Beeck Center leader reflects on ‘bipartisan opportunity’ for government efficiency

July 4 meant the end of the Department of Government Efficiency, as the federal initiative launched by President Donald Trump’s January 2025 executive order officially sunset.

DOGE says it has saved billions of dollars, but its record is unclear. What is clear is that the initiative caused chaos at various federal agencies with layoffs and closings, as well as efforts to use technology to solve vexing problems that led to many privacy concerns.

Against that backdrop, and as many agencies start hiring again, state governments also experimented with their own efficiency efforts. Some were high profile, like in Iowa, Florida and Texas, which set up their programs to try and imitate the federal version.

But other states have looked to rethink efficiency and instead define it as the effective delivery of government services. One such state, Arizona, talked up its planned monetary savings but also framed the efficiency effort as a way to rethink certain business processes.

At the forefront of thinking about government efficiency is Georgetown University’s Beeck Center for Social Impact and Innovation, led by Executive Director Lynn Overmann, a veteran of the U.S. Digital Service — which was repurposed to house DOGE via executive order — as well as other federal agencies. She also spent time as a public defender.

In the run-up to GovExec’s Government Efficiency Summit this week in Washington, D.C., Route Fifty caught up with Overmann to discuss where government efficiency goes from here, and how the trend is unfolding at the state level.

This interview has been edited for brevity and clarity.

ROUTE FIFTY: What does efficient government mean to you?

LYNN OVERMANN: I've worked both in the federal government and with state and local governments, so have a pretty broad perspective on it. But at the end of the day, efficient government means that government meets the needs of its constituents with the least amount of burden, both for the constituents and for the government workers that are delivering these services.

ROUTE FIFTY: Days after the end of DOGE, as you reflect back on that, what did you think of it, its aims and its outcomes?

OVERMANN: What was really fascinating to me observing DOGE is that it felt like it was repeating some of the lessons we learned when we first launched the U.S. Digital Service after the Healthcare.gov crash, which was this deep belief that the technology itself was the problem and that therefore technology was the solution. What we learned over time at the federal level — and I think it's being implemented in a far more innovative and effective way at the state level at this stage — is that it's not the technology. You need to understand who your users are and what their needs are, and then you need to build the process, the operations, the regulations, all the things that go around the technology to understand what is standing between the individual who needs the support and the service that's being provided, and then the technology often would come in at the back end. 

It was really interesting to watch DOGE come into it with, frankly, a much stronger remit than we had. They were given a lot more air cover, they were given a lot more executive support, they were sent into agencies, and the agencies were told that they had to work with DOGE. Even with that, I think they ran into the same barriers that we did, which is, it’s not a matter of building a snazzy new technology product, it is really deeply understanding how government works and trying to make it better. That is a very different process than building a private sector technology tool.

ROUTE FIFTY: What's happening out there in state government?

OVERMANN: I would name a few states in particular: Maryland, New Jersey, Pennsylvania and Colorado are really doing some really interesting stuff. Colorado, I think, is a particularly interesting example, because their digital service team started not long after the U.S. Digital Service team and actually took some of the alumni from the original USDS and the original White House Office of Science and Technology Policy to launch that team, so they've been around about the same amount of time. 

What the digital service team has evolved into under their new leadership is they've recognized that a digital service team on its own, even if it's highly empowered, is not as effective as an overarching technology shop, so have really stepped into the chief information officer shop, which brings with it a lot more responsibility, a lot more authority and a lot more capacity. It’s trying to transform the CIO’s office from a more traditional technology shop into a delivery shop. If I was to name one thing that's really happening at the state level that I think could be an example that the feds should follow at some point, it is recognizing that the goal is delivery, the goal is not necessarily a digital service or a technology tool.

I think we're seeing that in Maryland as well, where they have built out a really strong team and they've embedded folks in the key parts of the state government that not only build technology in partnership with agencies, but also have a lot of carrot and stick authority over the agencies themselves as they're doing procurement. They’re trying to wrap their arms around procurements to ensure not just that they've got a strong internal team that is capable of building a tool, but recognizing that a lot of government technology really comes from procurement, making sure that the procurement is well crafted, that there's a really clear sense of what the tool is that is needed, and then managing the vendors to ensure that it's actually delivered to the public.

ROUTE FIFTY: I want to ask about the goal being delivery. It must be hard as a government, because it must be very tempting to just bolt in a new technology and say, ‘Alright guys, we're done.’ How do you change that? Is that changing a mindset?

OVERMANN: It's both changing a mindset and changing an approach. One of the things that we've talked about for a couple of years now is something called the product model, which does tie back to procurement. It’s everybody's favorite thorny topic that is a challenge for everyone, but is recognizing that buying technology is not the same thing as buying a school bus, for example. The point of the technology is that it is responsive to the emerging needs of the program, and so technology is an ongoing piece of work. It is not a one-time procurement, and I think a lot of the procurements that we've seen — there's a number of reasons that procurements tend to go sideways — but one of them is this belief that you can name what you need the technology to do, you can buy the technology, you can deploy the technology, and that's the end.

I do think the mindset shift that we're starting to see, where delivery and outcomes are the goal, is an ongoing recognition that there needs to be ongoing work. User needs are going to change over time, technology is going to change over time, programs are going to change over time, funding is going to change over time, and what you really need is an infrastructure in place within government. A lot of it comes down to internal talent and capacity, who are capable of understanding in an ongoing way how those needs have changed, and how we should adjust the technology to be responsive in an ongoing way. It's not the way the government typically operates, and it's going to come from change management. That's why it's going to be so fascinating to watch these states that are pioneering in this space move to a goal, a new north star of, this is what they’re aiming for, and how do you bring these government agencies along to actually deliver it?

ROUTE FIFTY: Just on the procurement point, many states talk about having a request for proposals that's out of date by the time it comes into effect. How do we fix that?

OVERMANN: You need to have people on the inside who understand technology well enough to say, “What should these RFPs hold? Do we need an external vendor, or could we actually feasibly build this ourselves?” It's not always an obvious answer. I'm not ever a solid, "You must build," or "You must buy.” I think it depends on what you're delivering and what technology you already have in house, and then if you are going the vendor route, making sure that what you're asking them to do is deliver quickly to demonstrate that they are capable of building and delivering this technology, and then testing it.

It's really funny, we've seen over time the word “agile” has just gotten attached to a bunch of procurements, and, when you look at it, the actual process remains almost the exact same, which is you spend multiple years gathering requirements from different programs, you put them all into one incomprehensible multi-hundred-page document, you get two to three large vendors who respond because they're the only ones who have the capacity to actually dig through all of it and provide a response, and then by the time the project launches you can be two to five years into this process. Your requirements are out of date, and that's not even saying the amount of time it takes for the vendors to then build tools, because you've tried to capture everything all at once in one place. We're still talking four to five years for the tool itself to get developed, or for the system to be deployed, so you're anywhere from five to seven years into a product by the time you get to a point that you're starting to use it. The way technology is moving these days, that is just woefully out of date.

ROUTE FIFTY: Is it also difficult to do any modernization or efficiency project when you've got these old systems that can't break?

OVERMANN: There's been a really interesting project that emerged from the government of Alberta in Canada. They had a bunch of agency systems that were at the end of life, and they were facing a multi-year procurement, I think it was a $50 million RFP, with at best new systems in place in three or five years. What they did — this was a thoughtful use of AI — is the Alberta team had an empowered technology lead who actually understood what agencies’ technology needs were. He had a team underneath him that knew how to actually deploy using AI, so it wasn't just vibe coding to say, “Update my system.” It was, “What is the system architecture we need in place? How are we going to combine systems so that we are reducing redundancy? How are we understanding and meeting the agency's needs, and then deploying his team to use AI to build modularly?” 

They were able to modernize multiple of their systems in just eight months, and I think for under a million dollars. That’s an internally empowered team that deeply understands their users' needs, and, in this example, it was an internal government need using modern technology to build things quickly. I think that’s a path that more governments should be pursuing in the future.

ROUTE FIFTY: You’ve talked about change management and having folks on the inside who care about this stuff. How do you go about getting people on board?

OVERMANN: Having clearly defined outcomes and goals. The vast majority of people who work for government are there because they truly believe in the government's mission. Every career staffer I've ever worked with on any technology project, if you can show them that what you are suggesting that they do is different from what they're currently doing and will actually get the job done better and faster, they'll do it. The disconnect is skepticism that the delivery will actually happen because folks have seen lots of great ideas come and go that don't actually work, and I think that's a totally fair concern around whether it’s a flash in the pan or a sustainable approach.

One thing that we learned over time in the federal government certainly is bringing team members in, whether it's programmatic team members or whether it's traditional technology IT shops, and co-designing with them, making them feel as if they're a part of the project from the outset. Nothing gets people on board more quickly than success, and they know the problems way better than we do. Oftentimes, it's not that they don't know what better looks like, it's that they haven't been able to do it with the tools that they've been given. If you bring them in as the experts, and then you help them get where they know they want to be, that will handle a lot of it. If you just parachute technologists in, they come in, they don't actually understand what the delivery challenges are, and they say, “We're going to just deploy an AI tool to solve this in two days that you've been working on for 10 years.” It's not successful, and you're not having a lot of proponents for it when you walk out the door.

ROUTE FIFTY: What's next? What are you going to be paying attention to in this government efficiency space?

OVERMANN: I actually think there's a really interesting bipartisan opportunity coming down the pike. There's a ton of upcoming gubernatorial elections, and many governors who will be changing, either due to term limits or folks just not running again. There's going to be a whole new crop of governors at the state level. There's going to be a whole new crop of state legislators, and they're staring down a tremendous amount of opportunity with emerging AI and other technology opportunities, and they're staring down a huge implementation challenge, if you're just looking at things like H.R. 1.

The fascinating thing to me about state governments that I think is just so different from federal is so many states have balanced budget requirements, and they're forced to make trade-offs that the feds just aren't. Having an empowered, technologically savvy internal workforce and turning them loose on these projects and problems is probably going to generate a whole lot of new examples of how you can actually deliver government in a more efficient way, because you're going to have to in a more constrained environment, but you don't have a choice to not deliver these services.

I think there's going to be a ton of really compelling examples emerging from state governments over the next six to 12 months, and I think that's going to happen in both red and blue states, and I think that's a great opportunity to point out that government should work well, regardless of what your exact philosophy is on what government should be doing.

ROUTE FIFTY: Anything else you’d like to share?

OVERMANN: The civic tech ecosystem has also evolved in parallel with government work, and has gotten stronger and more robust with a lot of the lessons that we learned at the federal level. Us at the Beck Center, Code for America, U.S. Digital Response and others are full of folks who have been government practitioners working on these issues, and provide a really helpful outside space for government partners. We convene our government partners, and we give them a safe and trusted space to share their own problems with each other to shortcut. One of the things we've seen over and over in government is they often have the same challenge, but because they're not necessarily connected, they'll solve the problem themselves instead of understanding that the problem has already been solved.