Can Government Become More Scientific? D.C. Says Yes.

Let’s start with your elevator pitch for this work. What are you doing and why?

Broadly speaking, the idea is that whenever we’re doing something important, or creating a new program or policy, that we need to embed insights from science and research in the design, bring the users into the design process, and embed as rigorous an evaluation as possible into it. We may not succeed every time, but we’re going to know whether we’ve succeeded or not. That’s going to allow us to make better decisions going forward about how we serve the residents we care about.

Could you give me an example that illustrates your approach?

I have really enjoyed leading our work on the D.C. Flex program. Homelessness and housing instability have been a challenge for D.C., like many cities. When several of our district agencies shared an idea for a new rental subsidy that gave families more autonomy and would potentially allow them to serve more families, we helped them refine the design of the program. As we were designing the enrollment process for the program, we asked families at a local homelessness prevention services provider to test out the online application. They told us where they were getting stuck and pointed out aspects of the program or process that they didn’t understand. We then used that to refine our design. When we rolled out the program as a pilot to 125 families, we built in a randomized evaluation that allowed us to both fairly distribute spots in the program and measure its outcomes accurately.

Did we get it right when we designed the program? We’ll see this winter when we have the results from the first year. We hope they’re positive, but if not, we’ll know that we may need to adjust the program model or try something new to prevent homelessness.

If you think back to two years ago, what struck you about this opportunity when you walked into the office on your first day?

There didn’t seem to be a better place to try this work in the U.S. than right here in D.C. Government. Jenny Reed, D.C.’s Director of the Office of Budget and Performance Management, created my position out of one of her own existing ones. She took a relatively high-level position and said, ​“I want this person to build out The Lab @ D.C. and figure out how to make it work in D.C. government.” I really had only one job when I started and that was to work with David Yokum (who would become the Lab’s founding Director) to figure out how we do this.

There are also things structurally about the district government that are really exciting for this sort of work. You are much closer to the ground. Innovation can happen much faster here. If we’re trying to do something, whether it’s figure out how to better collect litter, administer a rental program, or build stronger police-community relations, you are much closer to the people that you’re serving and the people that are making decisions. Whereas the federal government or other places may have many, many layers of things that need to be approved, in D.C. government you often only need to convince three or four people that something is a good idea, and then it can actually happen and you can actually see it out in the world.

Research can pose some political risks because oftentimes, studies show that a program doesn’t work. Why is the D.C. government so committed to research and evaluation?

It really starts at the top. Even before the idea for the Lab formally existed, Mayor Bowser had a commitment to hiring people who really care about answering questions as accurately as possible. That came through really directly in some of the people that she brought on from the start of her first term. Rashad Young, who serves as City Administrator, was at the time — and still is really — just someone who loves digging into data, both budget data and performance data, and is committed to using that to guide decision-making in the district. I think he views The Lab @ D.C. as another tool in his kit to better serve residents and make better decisions about how we expend our scarce resources.

Jenny Reed, who I mentioned earlier, is someone who really cares about answering questions as accurately as possible. As budget director, she’s very interested in how we’re spending our money. We’re working to be able to provide input to those discussions and to be able to say, ​“If you do this thing, chances are it’s going to be effective.” Or ​“If you do this other thing, we’re going to be able to tell in a timeline that works for governing what we’re getting out of it.”

So it is both having a commitment to answering questions accurately and treating our data as an asset in how we serve residents. Then it’s also bringing in the people that both value this evidence-building in government and know how to get it done, which I think has been really important to us being able to get to this point.

In the years since the lab launched, ​“evidence-based policy” has become a buzzword among policymakers. Is it a trend or the future of the public sector?

I certainly hope that it is the future of the public sector, and I think we have some good examples of how it can work here in D.C. But it’s important to hone in on what we mean by ​“evidence-based policy.” 

When a lot of people think about evidence-based policy, they may tend to think about it as we know what works, we have a lot of evidence on it, and therefore we should just do it in whatever jurisdiction needs it. But what we found over time is that the best way we can do this work is to not so much focus on what’s been done elsewhere, but to think about what’s going on here in D.C., what would work best in D.C., and how we can embed science into whatever we’re doing.

We’re not as concerned about what worked elsewhere in Miami or Chicago or Houston. That should inform our work, but to me, evidence-based policy is thinking of evidence as we design policies and programs. We’re going to do the things that give us the best shot at serving residents, but we’re going to embed as rigorous a test as possible into what we’re doing to learn from it. The most important question is: ​“Does this work for the residents that we are trying to serve?”

If we can start to think of evidence-based policy not as something new in government, but as a different way of achieving the same goals that we already have, I think that will be really important to the long-term development of the field.

Have public expectations for government changed over the past 10, 20, or 30 years as research practices have become more sophisticated and we have gained more access to data and information?

I actually don’t think that they have. If you were to ask your average resident in D.C., ​“How should the district government approach a problem?” They would say, ​“You should look at what’s going on, listen to the residents, figure out if there are other things out there that have worked. Then you should do it, then figure out if that’s working or not, and if it is, continue doing it. If it’s not, do something else.” 

I think that’s how people think that government works and how it should work. I would agree with them.

What’s different is that we haven’t had as many of the tools to actually inform that question of, ​“Is it working or not?” And we haven’t been able to do it in a way that allows us to sync up government actions with results from rigorous research. So, it’s not that governments don’t care about results, it’s that it’s actually been very hard historically to get rigorous, scientific results in time to make a decision.

I don’t think of what we’re doing in D.C. as something that’s groundbreaking, or new, or anything like that. We’re just trying to add in a few additional tools and focus a little more tightly on the timeline for decision-making and how we are serving residents. If we do that, we’re going to meet residents’ needs in a way that’s much more aligned with how they think government should operate or how they think it does operate.

When you say tools, I assume you mean data collection and analytical tools. How do you build new data infrastructure at scale so that other cities integrate research and evidence into their decision-making?

The opportunity that we have and cities have is that we’re really starting to realize the value of the data that we’re collecting and that it doesn’t just need to be used for operational purposes. But I actually think of the ​“tools” as being more the people that we get involved who are able to use the data that we’re already producing in ways that can improve outcomes for residents. The bigger thing is spreading a scientific approach throughout D.C. government, and that happens through people.

We have The Lab @ DC, which is centrally located in the Office of the City Administrator and where our core staff work, but many people who are involved in the Lab are actually from different agencies. That includes the Metropolitan Police Department and the Department of Human Services. These folks are learning things about our scientific work and incorporating it into the work of their agencies. As a whole, this collaboration means that we have much more evidence-building and evidence-based decision-making happening all throughout D.C. government.

It’s really this spread of knowledge, tools, and techniques that I think is important for the development of evidence-based policy. The actual infrastructure, for data and other things, can be built off of that foundation.