Interview transcript:

 

Terry Gerton You have done some recent research that connects in very detailed ways the accuracy of census data to federal funding programs. Walk us through the high points and what’s at stake now as we look to 2030.

Sean Moulton Everyone knows that every 10 years we do a full census of the entire population here in the United States. And we’ve done it really since the founding of the country. But it’s not just an academic exercise just to figure out how many people are in each state or anything like that. We use that data in a very robust way. One of those ways is helping to guide our federal funding. What we’ve been looking into is, how much funding are we talking about that gets guided by the census? And what we found is 371 federal assistance programs that we can connect to census data in terms of guidance. And it gets guided in a number of ways. There’s some very simple ways where the census data can be, say, an on-off switch. The easiest example of this is funds maybe going to a rural area, or funds for an urban area. Urban-rural designation is entirely based on the number of people you have. You don’t have many people, population density is low, then you’re rural. That’s all there is to it. And so that census data getting that accurate can turn on or off money going to those types of areas. And there’s formulas where different pieces of census data go into an exact formula that figures out how much your area, your state, your county, your city or something like that might get. If you get the census data wrong, it could impact how much money’s coming to your area. A third area is a little bit more nebulous, but it’s definitely something we can track. Some programs accept applications and they can score and evaluate those applications on a variety of criteria, but they’re always transparent about it. And sometimes census data can come into play. Maybe the program is really geared and they want to help lower income areas or areas with historically disadvantaged communities. And so census data can be used to determine that, and your application might get extra points. And then the last way is for some of our loan programs, census data can even influence the interest rate that you might have to pay back. So it can affect how much money gets out and then how much you have to paid back. And these 371 programs, they accounted for $2.2 trillion in a single fiscal year, just one year.

Terry Gerton Was there anything about the 2020 Census or recent funding formulas that raised flags for you, that you want to make sure get addressed before we get to 2030?

Sean Moulton Every census, problems happen. It’s a huge endeavor, trying to count everyone in the country, at the same time, exactly where they are. We always have errors. But 2020 was one of the first years we did a lot more digital records. We were using what’s called administrative records to try and fill in some gaps from non-responses. And so we really need to address those. There are also a number of states that had statistically significant undercounts or overcounts, and those are particularly troubling. We need states and locations, especially in the areas that had previous undercounts, to make more of an effort in the run up to 2030 to make sure they get the count right.

Terry Gerton I’m speaking with Sean Moulton. He’s a senior policy analyst at the Project on Government Oversight. Let’s dig into those undercounts a little bit. Are there communities that are most vulnerable to being undercounted? And when they are undercounted, what is the impact?

Sean Moulton There are, and Census Bureau knows this and has made efforts over the years to do better outreach to what they call hard-to-count communities, or historically hard-to-count, communities. And these are lower income communities, because of the digital divide; these are rural communities; these are renters. Children are hard to count for some reason; even though the parents are filling out the forms, they might not include their children for some reason. Maybe they don’t understand it applies to everybody. Non-English speakers, not primary English speakers, sometimes they don’t understand the forms or understand the necessity to respond. So there’s a lot of different groups that are harder than the average citizen, we’ll say, to get those responses back from. This is where states and cities and counties can do a better job of reaching out and making sure their community members know the importance of the census, not just as a legal activity, but as something that helps the community and then responds.

Terry Gerton The census is supposed to count every single person, right? Citizens and non-citizens. We had the addition of a citizenship question in 2020. Certainly we’ve had a lot of focus under the Trump administration on citizenship. What impact do you think that’s going to have leading into the 2030 count?

Sean Moulton So in 2020, we had an attempt to add a citizenship question, and it went all the way to the Supreme Court and they tossed it out on a procedural issue. They said, it’s a question you can ask. It’s been in the census before, but they did it in the wrong way, their process was wrong, flawed. So we may be seeing another fight over that. The real problem with the citizenship question is there’s not much evidence that it’s going to give us anything of importance. More importantly, a lot of what we use the census for, it doesn’t matter if you’re a citizen. The funding for hospitals or healthcare or roads — the roads don’t care if you are a citizen and driving on them, or if you a non-citizen and driving them, and we need to repair the roads based on the wear and tear and how many people are there. The same for mass transit and other things. We’re funding for everyone. And so, if we try and narrow our ask to citizens, we’re going to get our allocations of funds wrong, and citizens will then be also penalized. They’ll have roads that aren’t being repaired fast enough and they’ll have problems getting into emergency rooms and what have you. Citizens will also be affected because they will encounter the problems that the low funding leads to: poor maintenance on the roads, longer wait times for their health care. And so even though they may think this is about citizens/non-citizens, everyone’s affected when the funding gets impacted.

Terry Gerton Do you have a sense that members of Congress understand this connection? I mean, at the core, one of their jobs is to bring home money to their districts. If the census count is accurate, the better their funding will be. And yet, do you think they really understand the importance of the accuracy here?

Sean Moulton I don’t. You know, it is pretty buried. We had to do a lot of research to figure out the extent of this. And I can tell you that just a few years ago, the Census Bureau used to do a report somewhat along these lines, and their number was much, much lower. It’s only been in recent years that we’ve kind of expanded our understanding to realize just how important the Census Bureau numbers are in terms of guiding federal funds.

Terry Gerton Are there steps that Congress or the Census Bureau should take now to improve accuracy coming into the 2030 census? I mean, five years seems like a long way off, but Census is already getting ready.

Sean Moulton  A lot of people don’t realize there’s a lot that happens in between those 10 years, but right now probably one of the biggest things would be to get ready to participate in what’s called LUCA, Local Update of Census Addresses. And this is a process that the Census Bureau runs in the run-up to every decennial census where they reach out and they try and get participation of local officials — county, city, state — to update the addresses they have. And an interesting fact is, if the Census Bureau doesn’t have your address, then it doesn’t matter if you fill out the form or not. You can’t be counted. The address comes before the household’s response. And so if somehow you’re living in a recently refurbished apartment over a garage and the post office doesn’t have that address officially on file as a new residence, then you’re not going to get counted. And so we really need to update those addresses and keep them as up-to-date as possible because it’s the first step to getting the responses back.

Terry Gerton Just to wrap up sort of on a more systematic note, is having this much federal funding dependent on the census the best way to go forward? Are there other funding formulas that we should use? Maybe even, wrong, it is the best source of data that we have.

Sean Moulton It is. Obviously, there’s other funding formulas that get used; 371 is not the majority of federal programs out there. But when you’re talking about trying to assist individuals and households, then the census data really can help us find those households and say, inside a state or inside a city, how much should they get? And we’re going to use data to help drive and allocate those rather than simply dividing it up into one-fiftieth and every state gets that amount. It doesn’t make any sense. If one state needs more, it should get more, and the census data, while we’ve had our problems, is still a very accurate number based on getting a lot of the money allocated. We get some things wrong and we’re always trying to improve that, but it’s still an incredibly useful tool for the federal government and for private individuals. Corporations use a lot of census data to figure out where they’re going to put their next grocery store or what have you. That’s because it has proven to be such a reliable tool to help guide those kinds of decisions.

The post An accurate census shapes how billions flow to states and cities first appeared on Federal News Network.

Interview transcript:

Terry Gerton You participated in an annual report from the American Statistical Association, their latest assessment, which describes this system as being at risk. We’ll deal with some of the specifics here, but let’s start at the top. How risky and why does it matter?

Nancy Potok Well, we’re talking really about 13 primary statistical agencies that are decentralized and embedded in different cabinet departments or agencies. We also have about 100 statistical offices or units that are small statistical areas across different agencies that aren’t designated as the principal agencies. But we focused on the 13 agencies. To give a couple of examples of what I’m talking about, the Census Bureau is a statistical agency in the Department of Commerce. So is the Bureau of Economic Analysis, which puts out GDP numbers, for example. The Bureau of Labor Statistics is one of the agencies. They’re in the Department of Labor. We’ve got the Energy Information Administration in the Energy Department. National Center for Health Statistics, which is in CDC, which is in HHS. I hope everyone understands the acronyms. So there’s 13 of them, and they’re in all different agencies — Transportation, Social Security, Statistics of Income in IRS, etc. And there is a chief statistician whose job, actually in the Paperwork Reduction Act, is to coordinate the system and to help the system come up with standards and quality standard operating procedures, what you should be doing in terms of confidentiality of data and privacy protection. That position resides in OMB, the chief statistician. And then there’s an interagency council on statistical policy — that is, all of the statistical officials that come together periodically to try to coordinate.

Terry Gerton I really appreciate the clear perspective there. And you participated in an annual report from the American Statistical Association, their latest assessment, which describes this system as being at risk. We’ll deal with some of the specifics here, but let’s start at the top. How risky and why does it matter?

Nancy Potok It’s very risky, and we’ll delve into the details of why it is, but it matters because these statistical agencies are what I would call essential infrastructure. They produce information that policymakers depend on, and when that information is not trusted or when there are holes in the system, that can really affect decision-making at the local, state and federal levels. In fact, sometimes local and state officials are much stronger advocates for federal statistical data than we’re seeing right now in Congress, where we need some champions. There are so many critical decisions that are made based on federal statistical data. I mean, for example, the Census everybody knows about — an apportionment of Congress, and that’s in the Constitution. But also, recently the head of the Federal Reserve was talking about his concern about Bureau of Labor Statistics information because it feeds right into their decisions on interest rights. So all across the board, there’s really a lot of key decisions that affect peoples’ lives that are based on this information. We want it to be accurate, high-quality, objective, not politicized and readily available to people to use, and people trust it and know that it is protected.

Terry Gerton The report documents a number of pressures on the system. The first one is staffing. It says that most of these agencies have lost 20% to 30% of their staff in the last year. How are you seeing that play out in operations, in data collection and reliability? How does it show up?

Nancy Potok It shows up in very scary ways. I’ll start with the most extreme example. I do have to say in fairness, some of these staffing reductions are just what I would call collateral damage of what we have seen across all agencies in all areas for staffing reduction. But I’ll start with the Department of Education. So the Department of Education has a National Center for Education Statistics. It’s a relatively small agency, but they put together critical information about how we are doing with educational achievement, and they gather that from all the states. As part of the downsizing in the Department of Education, their staff — everybody was fired, and there were only three people left in that agency. And the head of the agency was fired. So not only did they not have a leader, but they had three people left to do this job. And a lot of what they’re doing is required by statute, and they’re just unable to fulfill that. That’s at one end. The other end, the largest statistical agency is the Census Bureau. The Census Bureau had several field operations that were scheduled. They’re actually special censuses that they do for localities that need to update their populations because they’re rapidly growing and they can’t wait 10 years. And they couldn’t do them, they had to cancel them because they lost staff. In between what we saw was a multiyear effort that the statistical agencies had gone out to really get data scientists and people who could work with AI and who just were up on all the latest techniques that statistical agencies are using. All those new hires got laid off when…people who were on probation, because they were new hires, all got laid off. Years of recruiting, because a lot of the statisticians and data scientists are going to work in the private sector for very high salaries, so getting them into government was a huge multiyear effort, and then boom, they were gone. So, the next generation, gone. And then lots of senior people who were really keeping things going and had the institutional knowledge, they took the retirement. They just left. So we have gaps at the early stages. We have gaps in the leadership across the board. And then in some agencies, we have no staff at all.

Terry Gerton Nancy Potok is the CEO of NAPx Consulting and former Chief Statistician of the United States. Nancy, you mentioned public trust at the beginning, and your report shows that trust in federal data dropped from 57% to 52% in just four months, from June to September. What’s driving that decline and what does it mean for evidence-based policy?

Nancy Potok We didn’t delve deeply in the surveys that we were doing, which were conducted by NORC at the University of Chicago as part of their AmeriSpeak ongoing surveys. We’d like to delve into more of why, so I would just be speculating. But I assume it’s everything that people are reading about the federal government and the way information is being used. So if you’re reading that information is being used for law enforcement, for example, that agencies are sharing data for law enforcement — in the public’s mind, it may be difficult to sort that out. They are not necessarily that familiar with all the protections around statistical data and that it cannot be used for anything other than a statistical purpose. So if they’re worried about information being shared and their privacy and confidentiality, that could also reflect on their willingness to participate in the surveys. And we’re seeing response rates drop, which affects the quality. So it’s a bad cycle. In addition, when the head of the Bureau of Labor Statistics was fired, we had words that came from the president saying how bad the statistics were and really impugning the quality. So if people are listening to the president, which they are, that would also affect their trust.

Terry Gerton The report mentions nine new recommendations, and there were other outstanding recommendations. If Congress could move forward on these, which would you want to see them take on first?

Nancy Potok I know a lot of people talk about the staff and the resources, and there’s sort of a baseline at which these agencies operate that needs to be met. I think in my mind, and in the minds of my colleagues on this study, the thing that we want to see most is a strategy with some muscle to it to modernize the system. A lot of people have talked about, should we centralize these 13 agencies into one Office of National Statistics, which most countries in the world have? We have this very decentralized system. That’s a topic that’s been around a long time. What we’re saying is, look, there’s many things that can be done to improve this system. It’s not just money. We need a really strong chief statistician. We need some statutes that will drive modernization and innovation, kind of like a technology innovation fund almost, where we bring together strong leadership and give them the resources they need to work with partners outside the federal government. We put in a very strong pitch for bringing in the private sector and academia to do some of this heavy duty research on how we modernize statistics, and then bring them back to the agencies who are kind of busy maintaining, and then give them the resources to implement it. So let’s see some money go outside the government for this research. Let that research be directed towards how we have better federal data and statistics, and then let’s implement. But we need a strategy for that, and really strong leadership. I think for some people, when they were looking at the report, they viewed it as, “oh, they’re just asking for more money, like all agencies,” and that is absolutely not the case. I want to be clear about that. We are advocating for coordination inside and outside the government, removing the barriers to that and putting enough money in to fund what a strategic approach would prioritize. Because they’re doing a lot of stuff they don’t need to do, but they don’t have the leverage to stop doing those things. They still have to do those. So we just want to see almost like a remake of the system, a rethinking.

Terry Gerton Where in Congress does that responsibility lie?

Nancy Potok Well, that’s the problem because we’ve got 13-plus agencies and they’re all in different subcommittees. That’s part of what we identified as an issue, is that first, there’s no enough oversight. Two, the agencies, many of them are buried deep in the bureaucracy and they never get to talk to Congress directly. And third, kind of in the days when it was what I would call really the heyday of federal statistics, when they were very strong and supportive, there were champions in Congress. A couple of members in the Senate and House really understood the value of data and worked across all of these many different appropriations subcommittees and oversight committees to really think about this as a whole. We do have what I would call data champions in Congress right now, but a lot of times they’re focused on open data and just AI in general and those types of things. They haven’t really understood the role of the statistical agencies in that greater data ecosystem. And so that’s the challenge — getting the focus on that, to understand this is basic infrastructure of our country. It’s super important. Somebody’s got to pay attention.

The post The data system behind key U.S. decisions is losing staff, funding and trust first appeared on Federal News Network.