How to Investigate the Federal Government

"Sometimes folks will not fully appreciate our authority... our attorneys will have to educate them"

Over the past month, we’ve interviewed former bigwigs at three of Congress’s major support agencies, the bodies that exist to make Congress smarter, faster, and more effective at its job. In this series, we’ve tackled the Congressional Research Service, an in-house think tank. We’ve covered the Congressional Budget Office and its key role in shaping each year’s federal budget. And we investigated the now-defunct Office of Technology Assessment to better understand why it died and what comes next.

To close out this series, today we interviewed the Government Accountability Agency’s Orice Williams Brown. Williams Brown is the Managing Director of GAO’s Office of Congressional Relations. She’s previously led the Financial Markets and Community Investments team, and has produced GAO reports on myriad topics, including financial regulatory oversight, flood insurance, and Treasury’s Troubled Asset Relief Program.

What You’ll Learn

• How does GAO spot waste in the federal government?

• What happens when agencies don’t comply with investigations?

• How does GAO deal with fraud?

• Do Congress and GAO ever clash?

---

According to a March GAO report, the federal government made a quarter of a trillion dollars of improper payments last year. What happened?

Improper payments cover a range of types. There are payments that shouldn't have been made, payments about which there are questions about whether the people who received them were actually eligible for them, and issues related to not having enough information on the recipients to determine if they were actually eligible to receive them.

So improper payments really highlight a whole host of potential control issues that could indicate more serious issues in a program or at an agency. 

This report from GAO comes out yearly. What other regular products does GAO put out?

GAO has a number of what we call “signature products” that we put out on an annual or biennial basis. The first one I would mention is our duplication and cost savings report.

This is a required report that we produce each year, and it identifies just what it says, duplication, overlap, and fragmentation across agencies and programs in the government. We also include cost savings and revenue enhancements that could yield benefits to the government if our recommendations are acted upon.

At one point I was one of the agency co-leads on these reports. But it's an agency-wide effort. We’ll have a particular team or executive run point on it, but they coordinate activities across the agency. 

How does that work get done, on a nuts and bolts level? Are you asking agencies for information and collating it? Or is there an investigative portion as well?

There is definitely an evaluative or audit portion of the work as well. When we first started this work, we were mandated to produce this report. In the very first report, we basically went back over our existing body of work and pulled out places where we had organically identified examples of duplication, overlap, and fragmentation. That was the starting basis for this series of reports.

Since then, we actually do work at the request of Congress. Sometimes there are mandates. There are provisions in law for us to look into possible duplication or fragmentation in a particular program. In other places, we will proactively do work to determine if, in fact, that is the case in a program.

Then we try to determine whether it's really duplication, which actually happens less frequently. Pure duplication refers to programs or activities doing the exact same thing and providing a benefit to the exact same population. The overlap and fragmentation are much more common in terms of programs that overlap.

You have a particular population that is eligible to receive benefits from a cross-section of programs. Many housing programs, for example, are fragmented because there are several programs, and an individual could be eligible for some combination of those government programs. We really pull together this information from a cross-section of activities. 

Are those determinations that GAO makes ever contentious? Will agencies push back and say, “No, you're getting us wrong. This is not overlap or fragmentation.” 

We do have a back-and-forth conversation with agencies. They want to make sure we have a clear understanding of the program, the population served by the program, and how it works. I think there is always a tension with duplication, overlap, and fragmentation. You have to be careful. The assumption can’t be that if there are 57 programs that do X, then you don't need 57 programs, and all of them should be eliminated except for one.

But it's often much more complicated than that because there will be nuances about what population is served. Programs often proliferate because an initial program isn't fully functioning as it was originally envisioned and intended.

A new program will be created to address the perceived missing component of the program that already exists. And then, ten years later, someone will say, “Oh, but it doesn't really focus on this particular subpopulation. So we need another program to specifically address that subpopulation.” I think that is how programs with overlapping or fragmented missions tend to proliferate over time. 

Can you give me an example of those cases? 

I can definitely see a pattern: We have housing programs that cover a wide range of people, but we don't have a program that's specifically dedicated to veterans. Veterans have specific housing needs, so we're going to create a program that targets veterans. And then we'll say, “There's a subset of veterans that are homeless. So we need a program that targets the homeless veteran population.” And then among the homeless veterans, there are female veterans, but there was no program that focused on homeless female veterans. “So we need a program that focuses on homeless veterans.” And that's how it plays out, across a whole host of programs. 

I imagine that sometimes the GAO comes knocking, and an agency sees you as the enemy, or as a threat to a particular program that they’re running. They worry GAO is going to come in and say, “Those guys are useless.”

Does that dynamic ever occur? 

I think the relationship has really evolved. When I started at GAO in 1990, I would say it was much more adversarial. Now, and we've always said this, but I think more agencies tend to believe us when we say, “we're from GAO and we're here to help.”

That really is how we approach it. We take seriously our mission: to support the Congress in exercising its constitutional responsibilities, but also to look for ways to increase the efficiency and effectiveness of government on behalf of taxpayers and the American people.

We enter a conversation with that spirit. We don't do gotcha audits. We are clear with the agency about our key questions. We ask them questions. We also review documents and conduct physical inspections. We have a closeout at the end of the engagement, where we lay out all of the facts as we understand them based on the evidence that we have collected. 

We also give agencies an opportunity to comment on our draft reports, and we include their comment letters in our final published product. So they have an opportunity to weigh in. I think that many times, they understand and see the value of GAO and appreciate the fact that we can actually help them achieve their goals and make government work better.

I think everyone wants to do the right thing by — most people want to do the right thing on behalf of taxpayers. Many of them see that GAO can be an effective partner in that space.

I cut you off earlier – will you lay out the other regular GAO reports?

We also produce a biennial report called our high-risk report. This report is produced by GAO at the beginning of each new Congress. We lay out the top areas that need the Congress's attention, areas that need transformation. They're at risk of fraud, waste, and abuse. And we have a list that we track over time. 

In the intervening time between each new report, we’re doing work to update the areas on the list. We rate them against a series of criteria: leadership commitment, capacity, action plan, monitoring and demonstrated progress are the five areas they are rated on. The goal is to work with the Congress and to get agencies to effect the necessary change for it to no longer be deemed high-risk. Sometimes the Congress actually needs to take action to address a particular area on the list. 

What areas consistently show up on those lists? 

We started the list in the 1990s, and a couple of areas have been on the list for the entire time. Cybersecurity continues to be an area of concern and focus for us, as the threats and the type of risk and exposure for government continue to evolve, but that is an example of an area that's been on the list for for a really long time and still warrants dedicated attention. 

Other areas on the list are contract management and acquisitions. This area has been a challenge and permeates across government. Many agencies struggle to ensure that the contracting process is executed, overseen, and carried out appropriately.

What about the inverse? Are there success stories where GAO has spotlighted an issue and Congress has cleaned it up? 

Yes, we actually issued a report a couple years ago that focused on areas to provide a way forward and we laid out the keys to success. We wanted to make it clear to agencies that there are success stories. We found leadership, commitment, capacity, having a clear plan of action, and monitoring are all ways to do that. 

There are a couple of areas that we have identified where agencies took action. They provided dedicated attention, and they were able to come off the list over time.

Can you talk through some of those areas for me? 

One was about “federal excess and underutilized properties.” Congress really focused on this area because the federal government owns a lot of property, and for years, we’ve been pointing out that the government didn't have a comprehensive sense of where the properties were, their condition, and whether it still needed all of them.

So Congress passed — and the president signed — two reform bills to address this issue. OMB also developed a strategic framework to help guide agencies’ real property management. This has resulted in a decrease in lease costs. And it also led us to narrow the scope of this high-risk area in 2011, and then again in 2021.

We estimate that focusing on this area resulted in $3.6 billion in savings and an additional $4.7 billion in lease cost avoidance through 2023. That's an example of an area that actually resulted in financial benefits because appropriate action was taken.

I will also highlight the monitoring area and supply chain management issues. For example, Congress established statutory requirements for the Department of Defense (DoD) to submit comprehensive plans for improved asset visibility monitoring.

DoD leaders subsequently developed a detailed corrective action plan to carry out that requirement and a process for ongoing monitoring. Those actions alone contributed nearly $3.7 billion in savings between FY 2006 and 2019. As a result, this area came off the high-risk list in 2019.

I'm looking at a report from GAO that says the GAO generated an average return on investment of $133 for every $1 spent by the agency over the past six years. It sounds like there's a case to be made that the GAO is the most dollars-to-donuts effective agency in the federal government. Would you claim that? 

I think we definitely work hard to do our part. We think we play a significant role, and we think the investment in GAO is a sound investment of taxpayer dollars.

You mentioned most people are trying to do the right thing and be responsible stewards, but what's the most shocking misuse of taxpayer money you've uncovered while at GAO? 

I would look at the period of the pandemic. GAO recently produced its first government-wide estimate of potential fraud, and, in the federal government, we've looked at the period of 2018 through 2022. We estimate that anywhere from a quarter to half a trillion dollars annually was spent fraudulently. 

You had fraud in the Paycheck Protection Program, the Economic Injury Disaster Loan program at the Small Business Administration, there was organized crime involved there. There was Unemployment Insurance fraud, state-level fraud. It’s just mind-boggling. 

I think part of it is driven by the amount of funds expended during that period, but also just not having effective controls in place on the front end as the programs were being stood up. Some agencies have the notion that you have to choose mission or fiscal responsibility. GAO's position is you can accomplish both. You can have basic controls in place and still execute the mission and get funds out to people who desperately need them.

We recently developed a framework for emergency programs, to help agencies think about having the controls they need in place before the emergency actually happens, so they aren't scrambling to do it in the moment. 

I imagine agencies say, “Look, our mandate was to get money into Americans' hands as quickly as possible. We did our best.” So what should they have done differently? 

I think there are a couple of basic things. One is just to put basic controls in place. A lot of programs relied on self-reported income. We've pushed agencies to make sure that agencies have some basic controls in place. One of the programs involved just pushing out some basic funds to a group of eligible taxpayers.

So they were given hundreds of dollars of relief. And we found that a lot of those dollars were pushed out to dead people. Just basic controls like checking the Social Security Administration's “death master file,” whichTreasury now has access to. So there are certain basics that we believe will be beneficial for agencies to be able to verify before they push out funds.

Which agencies weren't checking to make sure that the Social Security number wasn't a dead person? 

So this was actually the Treasury’s Economic Impact Payments: which basically provided payments to certain taxpayers who had paid taxes in a certain period, that's what they used as the trigger point. Some received a check in the mail for up to $1,400. This was supplemental assistance at the height of the pandemic. At the time, Treasury didn't have access to the full death master file. With just that basic information, they would have been able to avoid sending checks to dead people.

And many of those checks were cashed by other people? 

Yes. 

---

You’ve had a long career at the GAO. How has GAO's engagement with Congress changed over that time? 

GAO is an independent agency within the legislative branch. So we've always had a close relationship with Congress to provide it with fact-based, nonpartisan, objective, non-ideological information.

We work hard to remain the source of that independent, nonpartisan information. The composition of the work that we've done for Congress has evolved in my 30-plus years at GAO. When I started, we did a fairly large percentage of work under the authority of the Comptroller General, which means it was work that we self-initiated. 

I have seen that percentage change over time. Now, the self-initiated work that we actually start under the authority of the Comptroller General is probably around the 5-6% range, and much more of our work is requested through mandates from the Congress. Congress will include a provision in a law or a committee report for GAO to conduct a study in a particular area. The rest comes from requests from the chair and ranking members of committees. 

And what accounts for that shift towards mandated products? 

I think the shift was intentional on GAO's part because we want to make sure that we're doing the work that Congress needs us to do. That was a big part of the shift. 

It has continued to evolve. The National Defense Authorization Act (NDAA) generally includes over 100 mandates for us to do work on a range of areas. The appropriations bills will also include a whole host of mandates. Mandates are our number one priority. They move to the top of the list because they're grounded in law and usually have a deadline associated with them.

Without asking you to criticize Congress, are there mandated reports that are better uses of GAO's time? I'm assuming some mandates you get are less useful for Congress itself, relative to other mandates. 

We really do work closely with the Congress. Committees often contact us before a mandate appears, asking us to weigh in and inform the mandate.

Usually, when we get a mandate, we are aware of it. It doesn't come to us cold. It will happen every once in a while, but for the most part, we are part of the conversation around something being turned into a mandate. 

That's also the case for requests. GAO has a five-year strategic plan. We work with clients and other external parties to frame the plan, but our plan is informed by the needs and interests of Congress. So, when a committee is contemplating sending us a request, they will often reach out to us to make sure it is something that we have the authority to do.

From GAO’s perspective, what kinds of things are you looking for in a mandate? Are there timelines on which GAO works better, or things you push back on?

Yes. When we look at timelines, it's usually just around a scheduling issue. We’ll say, “Hey, it may make sense to give us more time to complete this mandate because we’ll need a little bit of time to start the work.”

When bills are dropped, we’ll also look at draft mandate language, and every so often, something will raise constitutional concerns. GAO really isn't the appropriate agency to execute a mandate because it's a core executive branch function, so it really shouldn’t be done by a legislative branch agency. 

What's an example of that? 

Occasionally, draft language will imply that GAO will be making a decision about a program, such as if Congress wanted GAO to opine on a grant as part of the approval process for a grant program. 

If we're then asked to oversee the program, that could encumber our independence. That's where an issue could come up: We're asked to perform an executive branch function that then encumbers our ability to do oversight of the program because we've been involved in the decision-making process. That would be an example of a role it would be inappropriate for GAO to play. 

Any other common places where GAO will push Congress to shape a mandate in a certain way?

We want to be clear about the scope. Occasionally draft legislation will ask us to “find all examples of X.” We’ll often ask: “Are there particular programs, particular aspects or examples?” We'll try to scope the work to something that is doable in the time frame that they have in mind for us to do it. 

You mentioned that GAO's relationship with agencies involves dialogue. Are there agencies that might be less than forthcoming with data, or where information is hard for GAO to obtain? 

[pause] It’s a process with agencies. There are agencies that require more conversations. It's not uncommon with new administrations that you may have people who join the government who are unfamiliar with GAO. We have broad access authority.

Sometimes folks who are unfamiliar with GAO and its access authority will not fully appreciate how broad our authority is. So our attorneys will have to educate them about our audit authority and what we actually have access to. It's not a question of whether we'll get the information; it's how hard we are going to have to work to get it.

For the most part, agencies understand that. But when we encounter issues, we have a process internally where we escalate them and try to work through them to ensure we're getting the information we need in a timely manner. 

Will you sketch out what GAO's authority is in these cases?

We have broad access to information. If documents have been created, we have access to them. We have folks who have appropriate levels of security clearance. If we're dealing with classified information, we have people who know how to handle it appropriately. If an agency has information that we need to carry out our oversight responsibilities, then we have access to that information. 

---

We've talked a lot about GAO’s strengths. Are there places where GAO knows it has blind spots?

We work really hard to build a multidisciplinary workforce. Most recently, GAO was asked to build out our strength in the science and tech space. We have always had scientists and technologists sprinkled throughout the agency. At the behest of Congress, we pulled them all into one new team called the Science, Technology Assessment and Analytics team. 

We have grown that team, increased the number of experts. We work hard to have the skills that we need to do the work that Congress asked of us because we recognize that GAO is expected to cover the waterfront when it comes to the federal government. So we need to have skills and experts that cut across all of the federal government. 

We work hard to make sure we have a workforce reflective of that, and to continue to get better at offering a variety of services. Sometimes it's a quick turnaround request where a committee or member is looking for some specific information and it needs it quickly, to the in-depth, deep dive reports that include recommendations and a lot of analysis. We're working to make sure that we serve both of those lanes. 

One area that we continue to work on is ensuring that we're getting Congress information when they need it. We're making sure our work lines up with the congressional calendar and also with the requesters of the work, recognizing that members in the House are up for reelection every two years. We have to be mindful that we're leaving enough daylight for those members to be able to act on the work that we're doing for them. 

It's interesting you mentioned the challenge of timeliness. As you probably know, one reason the Office of Technology Assessment (OTA) was attacked in the ‘90s was that it couldn’t deliver products on time. Congress said, “What's the point of an agency if it doesn't give us information when we need it?” It was an existential issue for OTA. 

Definitely, and we're mindful of this, which is why we have built out the different types of services. With the tech assessments that we do, we actually have developed a short-range tech assessment, where we do something in a shorter timeframe so we can get information to the Congress on an emerging technology within a specific time window, so it's not going to take a year or longer for them to get that information. 

[We covered the history of the OTA in depth last week:]

---

What things cause GAO to move slower than you might like? Is it culture? Tech? 

GAO has, by necessity and design, processes that we have to follow. That's how we ensure the consistency of our products and what we deliver to the Congress. They are absolutely necessary. I think we just have to make sure that we continue to find ways to make the processes as efficient as possible. 

So I think that's probably a piece of it. It's always worth reassessing: Does it still make sense to do things this way, given technological advances, and are there ways that we can continue to sharpen the saw in that space? 

We'll soon find out the results from last year's employee experience survey and the best places to work. For the last three years, we've been the number one among midsize agencies in terms of the best places to work.

That's a testament to the agency's mission and our ability to attract people who are dedicated to it and really want to see government work better. 

What gets GAO those rave reviews relative to other agencies? 

The mission gets us a large part of the way there. But the other thing is that when it comes to leadership commitment to people, GAO is top of the heap. It's the tone that's set at the top, and we are as committed to our people as we are to the mission, and we make sure our people know that.