Group Looking to Fix Broken Government

Aaron Lemon-Strauss spent more than a decade in the private sector, developing and managing large-scale technology products for education. Then the Free Application for Federal Student Aid broke down.

The platform, which is critical to unlocking billions in federal student aid, had been undergoing a major renovation for years. The project was far behind schedule, but the deadline for the new platform’s rollout had arrived and technical glitches were affecting millions of students.

“FAFSA is the front door to America’s largest single investment in the future. It’s the way that we support Americans going to college: a path to the middle class. And that front door was broken,” Lemon-Strauss explained. “The government needed help from some private sector technologists to help get it back on track.”

So for the next two years, that’s what he did. From the end of the Biden administration through the beginning of Trump 2.0, Lemon-Strauss served as executive director of the FAFSA program. When he joined the team, the Education Department didn’t even have a job description for a single engineer. By the time he left, the platform had been opened to students ahead of schedule on multiple occasions and the staff included 40 new technologists to ensure FAFSA could continue to operate successfully for years to come.

In August, Lemon-Strauss moved on to a new role as chief digital and AI officer at a nonprofit start-up called Recoding America. This time he’s working on an even bigger project—fixing the federal government.

Inside Higher Ed spoke with Lemon-Strauss to hear more about Recoding America, a bipartisan group named after a book written by co-founder Jennifer Pahlka, and how its staff plan to fix what they call a broken government-operating model in a fast-changing world.

The conversation, edited for length and clarity, follows.

Q: Why don’t we start plain and simple? What does Recoding America mean when it says it’s going to address the current operational breakdown and change how the government works?

A: Fundamentally, in a democracy, the will of the voters and the people needs to be translated into operating policies, programs and government that is consistent with the will of those people and the representatives they elect.

But there are a number of ways in which laws, regulations or executive orders don’t actually translate into government operations consistent with the intent and will that led to their passage. It could be because there was a cascade of rigidity that led a system to stray from the original intent of legislation as it gets implemented. It could also be because of how the government buys and develops technology—[it] means that it wanders off, takes too long or costs too much to carry out.

But in each case we come back to this question: How do we have a government that is capable of turning policies into active everyday government delivery?

Q: So how do you answer that question?

A: We have four main pillars. The first is that we need a government workforce that has the right people in the right roles. That means—especially with the technology disruption that’s coming—we need more people in government who have the technical competency to be able to support a government in charge of its own systems. At the moment, too much of how government functions is outsourced to vendors. That needs to be brought into government.

The second pillar is about rightsizing procedure. Over many years, a stack of procedures have been built up on top of each other that make it harder and harder for citizens and businesses to actually interact with the government and the goals of its policies. So, rightsizing that is important and has bipartisan support.

The third pillar is about purpose-fit digital infrastructure. The government needs to have digital tools, websites and infrastructure that deliver on the intent of the laws and policies to provide outcomes for citizens. Too often, the digital infrastructure that the government uses is outdated and isn’t able to change with the times.

And then the final pillar is about having incentives for those outcomes. At the moment, much of how government functions is incentivized not based on outcomes, but based on procedure or avoiding bad press. We need a government where the purpose of a program is to deliver a specific outcome to whomever it’s serving, not merely to check a box.

Q: So would you say your work with FAFSA and the way you hired tech-proficient staff members was a demonstration of the first pillar, putting the right people in the right roles?

A: That’s right. And it also helped support the purpose-fit digital infrastructure and created outcomes-based incentives.

Getting the right people in place is the foundation for everything. And so the Trump administration were huge supporters of getting the right people in place, which then allowed for operating based on outcomes.

The technologists we brought in were not interested in the job just for following the procedure—although, of course, we always follow procedure. They were interested in the job for the outcomes that it produced. The way that we were able to attract top-level technologists to government certainly wasn’t the money; it was the mission. If you can say to a midcareer person, “In this role, you will have the autonomy to change millions of people’s lives and open up a pathway to the middle class for millions of families,” they will absolutely take that job as long as you can deliver on giving them the autonomy to do it, which Secretary McMahon was entirely behind.

And that helped create a culture shift that pushed the rest of the staff, who are all very talented and smart and hardworking, to pull their head up from the day-to-day procedural minutiae and look at who are the real users we’re trying to serve and what are the outcomes we’re trying to drive.

Q: Further tapping into your work at the Education Department, what other examples of operational breakdown did you see in the world of higher education, aside from FAFSA?

A: I’ll give you a good one that the team [at the Departments of Education and Treasury are] making phenomenal progress on.

When a borrower has entered default on their student loan, it’s a terrible experience for the borrower—their credit is hit, there are often jobs and housing that are not available to them, and it’s very hard to get out of. But one of the most straightforward paths out of default is to rehabilitate the loan. The process of rehabilitation, however, involved faxing documents in, waiting weeks to hear back and then being given a payment plan that was based on a bunch of data that you may not have access to.

So the team recently stepped in and said, “This doesn’t make any sense. We’re going to take this from faxing and turn it into an online real-time process that’s straightforward.” But what the team has also done is treated that application for rehabilitation of your loan as an application for an income-driven repayment plan that continues after your rehabilitation.

That is an example of insight about user needs. It’s better for the Treasury. It’s better for the user. And it’s better for society.

Nothing says that we have to do that in statute or regulations. It’s just happening because the team is focused on the user and on user outcomes so that they can produce better experiences.

Q: Going back to the note of proper staffing: The Trump administration made a historic reduction in force at ED and a number of other agencies in early 2025. Trump said the cuts were to make the workforce more efficient; would you say they were effective? And what would Recoding America do differently?

A: The reason that we need the right people in government is to deliver outcomes for users. So, often I find that the biggest change driver on personnel in government is the need for more technologists, not a reduction of existing head count.

Now, there are of course places where—because a program has lived past its shelf life, or maybe has overhired—you could see an opportunity for efficiency. But the biggest lever for change and better outcomes for users is often the addition of technologists into government.

Importantly, this does not need to cost additional money to the taxpayer, because often, or always, the government is already paying for this level of technology oversight. It’s just paying for it through a vendor. So shifting from sending the money to a vendor to having it insourced with government employees is a crucial move that allows for better outcomes, better monitoring and, in the end, more efficiency.

Q: Many Americans question the safety of AI and the speed of its development. As the chief digital and AI officer at Recoding America, how will you use AI responsibly?

A: The responsible deployment of AI is perhaps the best possible argument for the government having internal technical expertise. I want to live in a world where the people making decisions about the deployment of AI, the data it has access to and the products that it builds are civil servants who are accountable through the democratic process—not a world where the government has outsourced that decision-making and implementation to a vendor. Because that vendor, while they may be well-intentioned, is not in the end accountable through the democratic process. I want that person to have the right technical sophistication so that I’ll trust that they’re balancing the right equities on this and coming to the right decision.

So, AI, if anything, accelerates the need for the government to have technical competence within the state, so that we get the benefits of AI while being aware of and ensuring that we ameliorate any risks that come from it.

Q: To close, your website talks about how there’s no better time for Recoding America than now. So let’s focus on that: Why is this the right moment to launch an organization like yours?

A: The fundamental model of government came out of a different era. It came out of the postwar era, and the government continues to operate based on that fundamental model—some would say similar to how education has also come out of a Carnegie credit-hour model that goes back quite a long way.

Fifteen or 20 years ago, the private sector had to reboot to be digital. Legacy companies all had to shift to being digital, or else they were going to go out of business. The government faces a similar moment. The private sector firms are moving ahead on improved user experiences, and that then raises the user’s expectation about what a government should be able to produce.

Citizens are willing to give the government a discount on user experience, but when the bar keeps getting raised on what you can expect to be able to do on your phone or through chatbots, users start to expect something similar from the government. So if the government is stuck in amber and unable to deliver, that becomes a crisis; a crisis is defined by when a group of citizens are able to make enough noise to their elected representatives and the media to make it worth paying attention to. And we’re going to start seeing more and more crises.

It is time for a new operating model for government, one that is still connected to the right ideals and outcomes and serving citizens but uses and benefits from what technology can deliver.