Tuesday, August 12, 2025

NPR vs. DOGE

 Today, we’re taking a look at a predecessor to DOGE: The Reinventing Government project (officially known as the NPR, for National Partnership for Reinventing Government). The NPR ran for almost the full duration of President Bill Clinton’s two terms, and led to the elimination of over 100 programs and over 250,000 federal jobs.

Both NPR and DOGE are case studies in a long history of government reform efforts — some more successful than others. Our guest is John Kamensky, who served as Vice President Al Gore's deputy for the National Performance Review (NPR) for eight years. Kamensky was colloquially known as “Mr. Checklist” for his work organizing the Reinventing Government initiative.

We discuss:

Did the NPR actually work?
What was the Board of Tea Experts?
Why was the federal government subsidizing mohair?
NPR made the federal workforce older. Was that bad?
What doesn’t Elon understand about the federal government?


As I understand it, the Reinventing Government initiative was the longest-running government reform project in American history.

Yeah, that's how I understand it. I’ve researched previous reform efforts from the past century and haven't seen anything that had that kind of endurance.

In broad strokes, would you call the initiative a success?

I would say yes, it was a success. It changed what government employees understood their jobs to be. The goal was to empower them to be able to get results, and they felt that in many cases. We conducted a survey and found that after eight years, 40% of the civil service understood what we were trying to get at — that’s enough to create some momentum.

Even in future administrations, people still talked about some of the initiative’s outcomes. There were about 100 different statutes that included bits and pieces of our recommendations, and a number of administrative changes, like designating the deputy secretaries for various departments as chief operating officers and having them meet government-wide as the President's Management Council. I believe that has continued until today, although I don't know what the current administration's approach has been.

You’ve flagged that government reform efforts like NPR and DOGE are a recurring feature of American policy. Can you place the NPR in the context of the history of government reform efforts?

.... There’s a long history of approaches to government reform: there were efforts under Johnson and Nixon, and the Grace Commission under Reagan. Typically, advisers outside of government came in to advise on best practices from the private sector, but the Reinventing Government initiative was different. We turned to civil servants and asked them what needed to be done to fix government operations, focusing on the “how” and not so much on what the government should be doing.

What was the policy conversation like in the late ‘80s and early ‘90s, before the Reinventing Government initiative?

This was toward the tail end of Reagan’s Reform ‘88, which were largely administrative reforms. In the late ‘80s under President George H.W. Bush, the focus shifted to improving government performance and financial management, which led to the passage of the Chief Financial Officers Act (CFO Act) in 1990. I was at the GAO at the time, and I was tapped to look at how to measure and improve government performance. I wound up looking at what was going on in other countries as examples. (...)

So you're at the GAO in the late ‘80s. How did you end up at the Reinventing Government initiative?

David Osborne and I had an ongoing conversation — he was fascinated by what other countries were doing, and I was fascinated by what he was finding in the states. In a previous book called Laboratories of Democracy, he did a case study of Bill Clinton’s governorship of Arkansas, so he was in the Clinton circle. During Clinton’s presidential campaign, David Osborne drafted a speech on how Clinton could reinvent the federal government. That speech was never made, but the ideas stuck.

When they organized the transition team, Al From of the Progressive Policy Institute was leading the government reform piece. They brought in John Sharp, who was running the Texas Performance Review, and it was more like DOGE. They went into the agencies and terrorized them to find savings. On David Osborne’s suggestion, they came to me to get the international perspective.

Then in early March of 1993, Clinton gave a press conference, and said the administration would create a National Performance Review. I called my counterpart at the OMB to ask what was going on, and he said, “Well, this is the Clinton administration. This is live and unrehearsed. We don't know what's going on.”

A few days later, I get a call from Elaine Kamarck, and she says, “I just started working for Vice President Gore. He was asked to do this reinventing government thing, and David Osborne says that I need to talk to you.” They called me, Bob Stone, who was doing the reinvention of the Defense Department, and Bob Knisely, who was over in the Department of Transportation.

I told them what was going on in other countries. Bob Stone was a sort of renegade in the Defense Department, a civilian in charge of the Department’s military installations. He oversaw a large team and was trying to delegate authority; rather than having them ask whether to repave a parking lot in San Diego, he said, “No, you know whether you need to repave it, don't ask me.” So he talked about how the DoD had screwed up. Bob Knisely had worked in a lot of different civilian agencies and saw parallels there.

Elaine suggested we tell the Vice President about it. The three of us met at Bob Stone's house to figure out what we would tell the Vice President, because this was our one chance as civil servants to say what needed to be done to fix the government. Clinton had actually done total quality management when he was the governor of Arkansas and wanted the Reinventing Government initiative to consist of civil servants, rather than turning to business executives as Reagan had.

Bob Stone had developed a set of principles for his model installation program and suggested we use those. He made this “gold card,” which had those principles and Clinton and Gore’s names on it. Over the years, we handed out those cards to around 100 civil servants, but the first gold card was made in the middle of a snowstorm at Bob's house to give to Gore at our meeting the next day. Bob Stone is a storyteller at heart, and he said that we really needed to convince him of how screwed up the government is. He had a box of his “toys”: examples of how idiotic things were in the Defense Department. We decided to take it with us to the White House. The vice president has an office over in the corner, which Gore called the Square Office as opposed to the Oval Office.

The next day, we went and sat down with him, Elaine, and the deputy director for management of OMB. After we each told our stories, Bob Stone brought out the box. He took out a can of spray paint and told a story: the Defense Department hired chemists with chemistry degrees and tasked them with making sure that each can of spray paint hadn’t expired. They had to sign a little form that's stuck to the side of each can to say that the paint is still useful.

Then he pulled out a steam trap, which is a metal valve that takes dirt out of steam lines in buildings. These each cost about $100, and when they start leaking steam, they leak about $50 worth of steam a week. But the procurement people wouldn't buy a replacement until they got a bulk order. That let them get them for $90, but it took about a year to get enough orders, and you lost $50 worth of steam a week in the meantime.

Gore found this idiotic, and Bob Stone suggested letting the engineers on the base buy small items using a credit card rather than going through the procurement system. At the end, Gore asked us to come on for six months to help. I’d had no idea this was a job interview. (...)

I knew that getting detailed from the legislative to the executive branch would not go over well at GAO. Sure enough, Senator Roth was on the floor of the Senate a few weeks later, saying that this is a violation of the separation of powers. So I resigned from GAO and was picked up in the executive branch.

A vice president had never been to most of those agencies in person, so it created enormous buzz. He went from the US Department of Agriculture to the Department of Housing and Urban Development (HUD) to the central courtyard at the Pentagon, and huge crowds of enthusiastic employees showed up. The Vice President was on stage with the department secretary to listen to these people, and they would talk about some of the most embarrassing things going on in their departments that the secretary, of course, knew nothing about. And the Vice President was sitting there taking notes. When we finished our draft report, Gore asked us to change it to reflect some of what he had heard and make recommendations based on those stories. (...)

What did you prioritize in those first six months, and what did you leave on the table? That first set of recommendations did not include moving around agencies or consolidating functions.

Gore had very few dictates on what we should or shouldn't do. One of them was, “I don't want any recommendations to study this more. I want recommendations to do something.” Another was, “Don't move boxes. Fix what's inside them.” That was really sage advice, because typical government reformers want to make an org chart look rational by moving things around, even though it’s really hard and costly to do.

The last major department that we created was the Department of Homeland Security, and that was more than 20 years ago. That was really painful to create, and it took years to actually become functional. Gore believed that we had good people trapped in bad systems, which was the total quality management, W. Edwards Deming-type thing. That was sort of the mantra.

Do you think that gave NPR a longer political runway than DOGE has had? DOGE has taken on some reorganization initiatives, which come with a much higher political cost. You engender a lot more opposition.

In the second round of reinvention in ‘96, after Newt Gingrich became Speaker and the House went Republican, the NPR was asked to do a new report and study. That one was focused more on what government should do, rather than how. And much like the second Hoover Commission, it didn't land very well. Agencies were eliminated, the Interstate Commerce Commission was eliminated. My favorite was the elimination of the Board of Tea Experts, which was created in the late 1800s. (...)

With DOGE, Elon Musk has focused on holding up really egregious examples of government spending. They’re often overhyped or taken out of context, but showing millions of Twitter users what they’re cutting has been a huge focus. How much did the Clinton-Gore administration tout those examples of waste in public?

We didn't do that during the first six months. Then we issued the main report and there was a big event on the South Lawn. After that, there were 38 accompanying collateral reports for each of the agencies and systems on how to implement the recommendations in the main report. Promoting the things that were screwed up wasn’t such a big deal, and the GAO was finding a lot of them. Like, there were about seven agricultural field offices in every county in the United States.

Really?

Yeah, five to seven Department of Agriculture field offices in each of approximately 3,000 counties, including Brooklyn. There were all of these different agencies, and the Department never co-located field offices. We found closing the offices really hard because members of Congress wanted to keep them. The story that we got was that there were so many field offices because President Lincoln had said that no field office should be further than one day’s horse ride from any farmer, so field offices were about 20 miles apart. (...)

Let me ask you about some of the more difficult parts of NPR. The initiative famously cut upwards of 400,000 civil servants over those eight years; it wasn't until the last five years or so that the number of full-time federal government employees approached that 1993 peak again. How did you figure out where to cut headcount and implement it?

One later criticism of NPR was that as the number of federal employees dropped, a lot of outside contractors and what you might call Beltway Bandits ended up doing a lot of that work instead. You had this kind of dark matter version of the federal workforce: it just wasn't on the books full-time.

This goes back to empowering employees, which was one of Bob Stone’s mantras for the project. There were too many overseers and people in mission support functions — one in three civil servants were middle managers in what we called “management controls” — procurement, budget, personnel, legal. There was a 1:7 employee-to-manager ratio. Bob Stone was inspired by the business writer Tom Peters, who said that the average in the private sector is 1:15. That overhead is closer to 15% rather than 30%, so we set those two metrics as targets.

We wanted to go from 700,000 employees down to 350,000: moving 100,000 to the front line for service and returning the savings for the other 250,000 to the government. We had an initial target of 252,000, but agencies wound up cutting even more, in part because of the downsizing in the Defense Department after the end of the Cold War. We were looking at transferring some functions, like cooks, groundskeepers, and security guards at Veterans Affairs (VA) hospitals, to the private sector so that the government doesn't have to manage this enormous workforce. We saw how other countries implemented new public management theory to let agencies focus on delivering their core service and not managing police and groundskeepers, so that was the concept.

It didn't pan out that way because when we asked the agencies to cut headquarters, headquarters cut the field. And OMB wasn't really bought into what we were doing to reshape the workforce, so they let the agencies do whatever they wanted to.

Why wasn’t OMB bought in? That seems like a classically OMB thing to focus on.


Part of it was that they didn't think that the Reinventing Government people were rational. We were not traditional.

Because you were head-in-the-clouds folks?

A bit of that, but also that we were trying to empower employees, and that wasn't their shtick. OMB was more into top-down control. They have changed over the years — in fact, one of NPR’s outcomes was changing OMB, so that instead of being just budget analysts, they became resource managers. It wasn't just about money, but about how agencies function, and looking at their management and regulations.

But this didn't roll out the way we had envisioned it. There were some cases where it did, like when Jim King, who was the head of OPM at the time, cut down HR regulations and got rid of the 10,000-page personnel manual that agencies were expected to abide by.

He also worked with the agencies to downsize their staffs. That resulted in some hiring freezes, so you wouldn’t have new cohorts fall in over the years. Because of that, a gap of experience emerged, especially in procurement, which cascaded over the next 20 years.

Is that the gap the Beltway Bandits grew to fill?

In some cases, they did, but there are some functions that you just can’t delegate, like signing authority on contracts or coming up with contract strategy. We were hoping agencies would simplify HR and the procurement rules, which would let them do with fewer staff. But Congress ate dessert first and cut the number of people without simplifying the rules.

Is that what we're seeing with DOGE, where they start with the headcount cuts and hope that the regulatory cuts come later?

No, they think AI will fix it all, but it will be hard to have AI in the national parks to help someone who sprained their ankle. But they should have engaged Congress on the changes to the systems. In this current administration, we're not seeing any interaction that I can see with Congress of changing the rules of the game.

With those hiring freezes, there was a shift in the demographic composition. Federal employees started to skew older, and the share of federal workers under 35 fell by about 10% over the eight years of the NPR.

When you combine that with the median GS grade, the median place on the pay scale in the federal government is higher now than it was 20 years ago. That makes it a lot harder to fill roles with junior talent, or bring people through the ranks from a young age.

What do you make of that effect of the NPR?

There's also a shift in the mix of work that's done. For example, during NPR, the HR processes and agencies were manual, and they have since become electronic. In many agencies, small-time procurements were manual, and you had triplicate forms that took weeks to complete, and the administrative cost of processing those forms was horrendous, so they moved to credit cards. Interestingly, the current administration has reneged on the use of credit cards, and it's costing the government money. In fact, when there was credit card abuse in some agency late in the Clinton administration, Congressman Pete Sessions put in a bill to stop the use of credit cards, but the Congressional Budget Office came back and said, “That'll cost you $100 million in rebates.”

That's amazing. The rebates were a substantial chunk of money.

Yeah. You have to balance risk with the ability to get stuff done. If you want no risk, it’ll be very expensive.

Right. The current administration wants to centralize procurement at GSA, and we've seen some moves in that direction. As you describe, there's a constant pendulum swing between pushing purchasing and management authority lower into the system, and then realizing the risk and pulling it back. Then, once it's centralized, you realize that you've limited your ability to trust the outer branches of the federal government.

There are pros and cons of centralizing. One of the things that we saw was that, at the time, the US government was the last vestige of Soviet central planning. We needed to devolve that. The government had a choice between different administrative service centers, which we call franchise funds, and those still exist today. There were six or eight different franchise funds created across the government, so the Department of Health and Human Services would be able to service contracts and HR for other agencies, which allowed for transparency about how much an administrative process costs and created some competition between agencies. If an agency head gets fed up with how their procurement contracts are done, they can go to another agency. By recentralizing, you will initially gain some efficiencies, which will turn into inefficiencies over a few years as it becomes a non-responsive monopoly. (..)

If you and I are talking again at the end of this administration, what are we likely to see as the effect of DOGE four years from now?

I think we're going to see agencies lose the capacity to do things that Americans assume just happen, and that they'll see enormous holes in the safety net of what government does and how it delivers. The Clinton-Gore administration was very concerned about losing citizen trust in government, which would make it hard to maintain a democracy. By the end of the administration, it had gone from something like 20% to 40%, and now it's really low again. With DOGE, I think it will fall even further.

Let’s say Elon really is stepping back, and imagine you get installed at DOGE. Everything that’s happened already is already baked in, and there are a bunch of headcount cuts that are not being rolled back, a bunch of attempts to centralize procurement, the implementation of these digital systems, etc.

How would you try to right the ship?


You can't put the toothpaste back in the tube. I don't think you can go back to where we were. The question is, “What will government look like next?” I think we’re going to have a period of huge turmoil in terms of service delivery, like no one answering the phones in the Social Security Administration, and that the current immigration enforcement efforts will lead to loss of trust in the immigration system. Will student loans still work? A lot will get broken down, and I think a lot of people are going to get hurt.

So we need to ask how to develop a new approach or system, and I have not yet come up with that answer. There is a group of people that seems to be trying to develop that, and that is where my hope lies. I think a lot of it will be a matter of using technology. We've broken things so far, but we haven't come up with a plan for how to replace or fix them in a way that makes a difference for people's lives.

by Santi Ruiz, Statecraft |  Read more:
Image: uncredited
[ed. So, in this administration young tech bros not burdened by history (or anything else apparently, including puberty), were put in charge with no clue - and continue to be embedded like viruses, sucking up classified data.]