In this episode, I talk with Denice Ross about the fragile state of federal data and her new work at DataIndex. We discuss how the removal, alteration, or disruption of federal data collections and data products affects research, policymaking, businesses, and everyday life. Denice explains the vision behind DataIndex and EssentialData, which monitor data risks, highlight data dependencies, and help users recognize how federal data benefit society. We also explore the roles of private-sector data, civic tech efforts, and public advocacy in building a more resilient national data infrastructure. It’s a fascinating look at what’s at stake when the country’s information backbone is no longer guaranteed.
Resources
Check out America’s Essential Data for more.
Guest Bio
Denice Ross is a Senior Fellow at the Federation of American Scientists, focused on building a more resilient national data infrastructure. Most recently, Denice served as the Deputy U.S. Chief Technology Officer and as the U.S. Chief Data Scientist, where she led the charge to use disaggregated data to drive better outcomes for all Americans. Denice’s 25-year career in using data to serve the public interest has spanned federal and local government, academia, and the nonprofit sector, plus domains ranging from climate to policing. She served as a Presidential Innovation Fellow for the U.S. Department of Energy and as Director of Enterprise Information for the City of New Orleans. Prior to government, Denice co-directed the non-profit data intermediary Data Center, where she collaborated with Brookings to track New Orleans’ recovery from Hurricane Katrina. She brought a data-driven approach to numerous post-Katrina community planning initiatives and co-founded the first new childcare center after the storm.
Different Ways to Listen to the Show!
New Ways to Support the Show!
With more than 250 guests and 10 seasons of episodes, the PolicyViz Podcast is one of the longest-running data visualization podcasts around. You can support the show by downloading and listening, following the work of my guests, and sharing the show with your networks. If you’re interested in financially supporting the show, you can sign up for my Patreon platform, make a one-time payment via PayPal, or shop one of the show’s sponsors.
Transcript
00:01.66
Jon
Hi Denise. Great to see you again.
00:03.58
Denice Ross
Hey, John, you too. Yep.
00:04.89
Jon
I mean, let’s see. I saw you what, like a couple of months ago at this USA facts event in DC. very excited to be like focusing on this challenge we have with federal data.
00:14.52
Denice Ross
yep
00:19.78
Jon
So let’s just get into it. So um a lot’s changed in the last like 12 months. um You left your position as US chief data scientists at the end, right?
00:33.83
Jon
the end of 2024.
00:36.13
Denice Ross
Mm-hmm.
00:36.95
Jon
What have you been what have you been up to since then last twelve months
00:42.26
Denice Ross
Yeah, well, um so a few weeks after the election and in 2024, group of what I’ll call federal data watchers, um and I convened in my dining room.
00:55.24
Denice Ross
And I took all the pictures off the wall, and we got out the Post-it notes, and we ah you know we we looked back to the lessons from the first Trump administration and what happened
00:59.00
Jon
Yeah.
01:07.06
Denice Ross
with with data in that administration and what we learned, especially around climate and environmental data. And we mapped out all of the things that could go wrong um and how we would know what was happening and what we might do to sort of head off the worst the worst impacts on the the nation’s capacity to to use data to inform all of the the things that we need to do as a modern society.
01:32.31
Denice Ross
and um yeah You know, it was it’s been a rough year from a data perspective.
01:38.00
Jon
Yeah. Yeah. yeah
01:39.84
Denice Ross
That’s no surprise. um and And what I’m doing now is ah the first working with a group of colleagues to i to to monitor changes to federal data sets.
01:54.42
Denice Ross
and ah and assess figure out how to assess the risk of a given data set for not continuing, and then looking for these policy windows where public input might have an outsized impact to protect the flow of a federal collection.
02:11.90
Denice Ross
um Often that manifests in like federal register notices, for example, and and places where data users aren’t typically spending their time, right?
02:21.21
Jon
Yeah.
02:21.50
Denice Ross
as As data users, we just take for granted that these data will continue flowing. um And so that it is, you know, I think 2026 is going to be a good year for people who use data to start thinking about how to also become data policy advocates at the same time.
02:38.52
Jon
Right.
02:39.45
Denice Ross
So we’re doing that work through a website called dataindex.us.
02:39.92
Jon
So,
02:42.94
Jon
mm-hmm.
02:43.93
Denice Ross
it’s um It’s pretty wonky, ah but um ah but we we have a newsletter.
02:46.04
Jon
ah
02:49.27
Denice Ross
And so about once a week, we’ll just send the highlights of what opportunities there are for public input around specific federal data sets.
02:57.82
Jon
And so how are you, um, doing that part of it? Cause I think one of the challenges data people have had is how to keep on top of say the federal register. It’s not like there’s an easily accessible newsletter that comes out with all the notices. So are you in your, the folks you’re working with just like mining that every week to see what’s popping up?
03:21.75
Denice Ross
and ah Yeah, we scraped the website.
03:23.45
Jon
ah
03:24.34
Denice Ross
um there’s so um And it’s it’s worth sort of stepping back and thinking about what we mean when we talk about federal data.
03:25.11
Jon
Yeah.
03:31.42
Denice Ross
we um It depends on what field you’re in, right?
03:32.10
Jon
hmm.
03:35.11
Denice Ross
Like I think you and i you know, um you know we we came of age professionally sort of working on on data about people.
03:35.67
Jon
Yeah. Yep.
03:41.62
Denice Ross
um And so data collected through forms and surveys that are subject to the Paperwork Reduction Act, which there’s a lot of derision around the Paperwork Reduction Act and how it makes it harder to serve the American people sometimes. But it also requires some transparency and accountability and engagement around changes to federal collections. That’s really important.
04:04.34
Jon
Yeah.
04:04.50
Denice Ross
um And so any any data set that is collected through a form or survey, um if a change is being made to it, it has to go through this formal process of public notice, and then there’s a comment period.
04:17.06
Jon
Right.
04:17.08
Denice Ross
um But there’s so many other types of data. For example, the satellite data, right? um The geospatial data, there’s there’s reference data, like the census tiger line files, um or the um ah the the um ah the FDA has a ah data set that provides what the generic drug equivalents for name brand drugs are.
04:23.56
Jon
Yeah.
04:29.61
Jon
Mm-hmm.
04:42.29
Jon
Mm-hmm. Mm-hmm.
04:43.03
Denice Ross
Right, so these are all these are all data sets. there’s There’s data sets that help us navigate the federal government. So there’s clinicaltrials.gov, for example, which that behind that is there’s a data set and an API that has the characteristics of clinical trials. So if, you know, you you’ve got a specific type of cancer, you can go to this federal data set and see what clinical trials are available. um And and then then of course, there’s all of this really rich geospatial data, like the the National Geodetic Survey that that gives us a single source of truth for latitude, longitude, and elevation, even as the earth is shifting and wobbling.
05:23.16
Jon
Yeah.
05:23.83
Denice Ross
um And that you know stuff like that’s really important. Like we don’t want every self-driving car company coming up with their own source of truth for lat, long, and elevation.
05:32.25
Jon
yeah
05:32.73
Denice Ross
Right? We want them to stay in their lanes.
05:33.34
Jon
Yeah. yeah yeah yeah yeah Yeah. Those cars should all agree. That’s right. that They should all agree.
05:38.81
Denice Ross
Yes.
05:39.70
Jon
Yeah.
05:39.96
Denice Ross
Yeah. And there’s, you there’s just something that only, only the federal government can do.
05:40.51
Jon
Yeah.
05:43.24
Denice Ross
um
05:43.83
Jon
Right, right. um So ah there’s a lot to talk about, but but let’s let’s stay on this topic. Like from your perspective, what can the individual data users, and that could be researchers or analysts or people who are just, you know, they’re just want to make a cool map of something.
06:03.54
Jon
In addition to sort of following the federal register, like what other actions can they, can can an individual take?
06:10.42
Denice Ross
Yeah, so here’s something really simple that we as um as sort of tool builders or map makers even policy analysts don’t always do a great job of. And that is citing your source explicitly.
06:25.59
Denice Ross
you might um You might have been following the last decennial census, for example, when um there’s a census behaviors and attitudes survey. And the the the Census Bureau asked members of the public if they thought that census data was unique.
06:43.70
Denice Ross
And one common answer was, no, you know’s there’s nothing unique about the census data because you can get it from Google.
06:53.21
Jon
ah
06:54.52
Denice Ross
Right?
06:54.81
Jon
Right. Yeah. Right. You just get it from Google.
06:56.33
Denice Ross
Similarly, yeah get it from Google.
06:57.22
Jon
Yeah.
06:58.67
Denice Ross
Similarly, if um you know ah ah you know a football coach might be using you know a weather app in their pocket to know when the heat index like crosses an unsafe threshold so he knows when to move football practice inside, um without the National Weather Service, that that app wouldn’t have the data it needed.
06:58.79
Jon
Yeah.
07:10.46
Jon
e
07:17.62
Jon
Yeah.
07:18.14
Denice Ross
um or or if you’re moving to a new town and you know, you’re using Zillow or Redfin or one of these real estate apps. And, you know, one of the factors that you’re considering is, ah you know, the quality of the schools in the neighborhood.
07:33.21
Jon
who
07:33.47
Denice Ross
That’s coming from the Department of Education, but it’s like about three levels removed.
07:38.14
Jon
Yeah.
07:38.87
Denice Ross
And, you know, and it’s not like these real estate apps are being like, here’s, you know, brought to you by federal data.
07:43.35
Jon
Yeah. Right, right, right, right.
07:45.40
Denice Ross
Right. They take all the credit.
07:45.63
Jon
Yeah.
07:46.52
Denice Ross
Right.
07:46.78
Jon
Take all the credit. Right.
07:47.83
Denice Ross
Yeah.
07:48.12
Jon
But they’ve ingested all these different federal sources. I mean, not only that, like the Zillow example is a good example, like all the maps that they build.
07:54.95
Denice Ross
Mm-hmm.
07:55.13
Jon
I mean, those are ingested using federal geospatial data.
07:55.29
Denice Ross
Mm-hmm.
08:00.25
Denice Ross
totally
08:01.40
Jon
the The other one that I’ve heard in talking to people, ah I think more geared towards researchers is to, um I guess, promote the quality of the data.
08:13.02
Jon
Like I think there’s like researchers tend to say, oh, this data, you know, i use this data, but it doesn’t have x Y, and Z, right? And so like I, you know, could only do this thing.
08:19.80
Denice Ross
m
08:22.17
Jon
But there’s a flip side to that, which is, but it has all this other stuff that I can get nowhere else.
08:26.62
Denice Ross
Right.
08:27.73
Jon
um And so I’ve heard a few people say, like, ah from particularly researchers, but I think it extends to anyone using data to just say, like, this is the highest quality data we have.
08:40.64
Jon
And without it, I couldn’t do any of this, even though there are lots of other things I wish I had.
08:46.01
Denice Ross
Right. and And that’s the interesting tension with federal data, right? um Another project I’ve been working on is ah documenting the ways that specific federal data sets benefit everyday American lives and livelihoods.
08:59.90
Denice Ross
And we and that that work manifests at EssentialData.us.
09:00.09
Jon
Mm-hmm. Mm-hmm.
09:03.86
Denice Ross
And we’re almost at 100 use cases, 100 examples of um of how federal data impact people’s lives. And when I talk to data users, like people who know data sets really well inside and out, ah and I ask them about their favorite federal data set, they almost always say like, I can’t live without it.
09:24.36
Denice Ross
It’s the best thing that we have. And then they say, but it’s got these shortcomings, like the, you know, it’s got too much latency and it doesn’t go into enough geographic detail.
09:30.83
Jon
Yeah.
09:32.24
Denice Ross
And I wish they would disaggregate by X, Y, and Z. um And so we all, you know, you’re you’re right.
09:35.16
Jon
yeah
09:38.00
Denice Ross
Like we’re all of two minds. Like we desperately need this data and it’s also not, not meeting the needs, especially of a rapidly changing world and society.
09:47.38
Jon
Yeah, so wanna look for a moment, look beyond, and I think we should pull back and give people sort of ah a background of where things are right now. But when you think about the next five, 10 years of the federal data system, where do you think, and we can even just sort of like carve out where we are right now and to sort of ignore it, but like, where is the where do where do you kind of wanna see the federal data system be in another decade?
10:18.23
Denice Ross
Yeah, well, we it’s clear that we need to build more resilience into the system.
10:22.64
Jon
Mm-hmm.
10:23.83
Denice Ross
And ah one ah the the Evidence Act that that President Trump signed in 2019, I think, was a really good first step there. um And you know when i um so I served in the Obama administration as a presidential innovation fellow and worked a lot on policing data and climate and environmental data. And then and then when I came back in the Biden administration,
10:48.92
Denice Ross
I was struck by how much more capacity the agencies had. you know most of The major agencies all had a chief data officer, there’s a CDO council, um there were um ah ah you know data scientist jobs, and there were hiring pools, and like so you know a lot of progress had been made in in building the capacity of of you know of agencies to to use data to to ostensibly drive better outcomes. And um and then in in the Biden administration, we took it to the next level.
11:23.32
Denice Ross
um My role as the chief data scientist was to build the data backbone so that we could disaggregate data, identify disparities, and then fix them.
11:34.45
Denice Ross
Right?
11:34.84
Jon
who
11:35.05
Denice Ross
Like if you don’t have the detail, you can’t fix the problems.
11:35.70
Jon
Right. Yeah. Right.
11:39.03
Denice Ross
um And and that was that was really important. And that that’s when we talk about disaggregating data, it it includes things like rural versus urban, you know just the small area geographies and the discrepancies that you see you know based on based on place.
11:53.35
Jon
Mm-hmm.
11:56.97
Denice Ross
It includes things like veteran status, whether somebody is a caregiver for an an injured an injured veteran or whether someone’s a military survivor.
12:03.55
Jon
Mm-hmm.
12:08.27
Denice Ross
you know there’s different um There’s different ways to slice and dice populations, race, gender, ethnicity, um income levels, access to a vehicle. like It really really depends on what the research question is.
12:18.76
Jon
Mm-hmm.
12:19.97
Denice Ross
And and in the in the Biden administration, OMB made really good methodological progress on updating, like modernizing the race and ethnicity categories.
12:33.26
Denice Ross
for disaggregation and also made progress on ah on some emergent ah best practices for collecting data about sexual orientation and gender identity.
12:43.74
Denice Ross
um And ah however, still… we still the system was still so highly fragmented. um And there were there were these weird divisions.
12:55.90
Denice Ross
And I’m sure you felt this in your career, like we treat our tabular data so differently than our geospatial data. Right?
13:03.58
Jon
Yeah, yeah, yeah, yeah.
13:03.90
Denice Ross
And like and that’s just, it’s silly. and the And the silos, you know, I’ve heard that there are there are, you know, like dozens of ways that the federal government collects data on rainfall, for example, precipitation.
13:17.67
Jon
Hmm.
13:18.55
Denice Ross
um And so like as ah as a local practitioner in when I worked for City Hall in New Orleans, it was really hard for me to even figure out like which federal data set is most relevant to us in terms of future planning for future storm surge and rainfall.
13:19.45
Jon
Hmm.
13:35.08
Jon
Right.
13:35.99
Denice Ross
um And and it shouldn’t it shouldn’t be that hard. It shouldn’t be so siloed. um
13:41.85
Jon
Right.
13:42.52
Denice Ross
and and and And going back to the Evidence Act, I think one thing that was really brilliant about it is there was a requirement in there that federal data stewards engage with the public about the value of the data and how it could be improved.
13:59.10
Denice Ross
what What we haven’t done a good job of is thinking across the federal silos.
14:04.09
Jon
Yeah.
14:05.05
Denice Ross
um So so that that there’s not that artificial distinction between data collected in specific different components across agencies. And really more of a focus on the users.
14:14.65
Jon
so so So what in your mind is the role of, um I guess, individuals, but also like private sector companies, private sector organizations in business?
14:31.86
Jon
I guess, improving or streamlining, i mean, all these things, I guess, improving sort of takes care of all it, but what is the role like the private sector and individuals in improving the the federal statistical system?
14:44.73
Denice Ross
Yeah, that’s a really good question. ah Sometimes what i i I zoom out a little bit and I think about building a more resilient national data infrastructure, thinking about um some points of resilience that have happened organically that maybe we should be more intentional about.
14:49.88
Jon
Yeah.
14:54.03
Jon
Mm-hmm.
15:01.27
Denice Ross
and One good example was with COVID, um how entities like the COVID Tracking Project and Hopkins stepped in to harvest the data from state and local public health agencies and harmonize it when the CDC sort of hadn’t hadn’t caught its groove yet with that collection.
15:08.92
Jon
Yep.
15:19.38
Jon
Yeah.
15:19.85
Denice Ross
And that that went on for about a year. um And more recently, last September, well, two Septembers ago, the Jeff Asher launched the Real-Time Crime Index,
15:34.05
Jon
Yeah.
15:34.17
Denice Ross
and And it similarly harvests data crime data directly from jurisdictions and gives a monthly s national estimate of crime. And um and that’s in that’s complementary to the official statistics that the FBI produces on an annual basis, right?
15:43.97
Jon
Yeah.
15:52.04
Denice Ross
but But the way the FBI does it is the year closes and then it takes them nine months to process the data to make sure that the quality is the level that it needs to be if it’s an official government statistic.
15:57.65
Jon
Yeah.
16:02.96
Denice Ross
um but in the But in the meantime, policymakers and the the public need need access to this information.
16:03.13
Jon
yeah
16:09.56
Jon
Yeah. then Yeah.
16:11.26
Denice Ross
Yeah, and so so i there’s um there’s there’s a ah real beauty in the innovation that can happen when you’re outside of government. And you don’t have you don’t have to worry about getting called before Congress to explain why you had to you know revise number, right?
16:19.89
Jon
Mm-hmm.
16:24.50
Jon
Yeah. Right. Right.
16:26.33
Denice Ross
you just You’ve got your method.
16:26.66
Jon
Yeah. Just do it.
16:28.01
Denice Ross
It’s like good enough.
16:28.81
Jon
Yeah.
16:29.05
Denice Ross
you’re gonna you know and And so that that tension between quality and timeliness, you know, having having some of these um sort of scrappier ah collections that are closer to the source can can help us
16:33.98
Jon
Yeah.
16:42.78
Denice Ross
build a more complete system where we’ve got the official data come coming with a little bit of a lag, but you know very high quality, um not as flexible in terms of like the you all of the data collected, but covering the core. And then you can have innovations, whether they’re more timely, that might add on new attributes, for example, um you know experiment with collecting data about non-fatal shootings, which until recently wasn’t in the federal standard.
17:10.30
Denice Ross
but seems super important for understanding violence in America.
17:10.42
Jon
Right.
17:14.01
Jon
Yeah. Do you have… Strong feelings or even a proposal on whether the future of the U S system is more like, uh, to, to, to sort of address these siloing issues is more like a statistics Canada type approach where it’s like, there is a single like statistical agency versus the way we, the U S has sort of set up now, which is a little bit more, I guess, you know, topic sort of specific.
17:46.30
Denice Ross
Yeah, my, um so what I would say is if we’re talking about data, i think statistical data is just one type of data.
17:50.94
Jon
Yeah.
17:55.31
Jon
Mm-hmm.
17:55.86
Denice Ross
And um it’s ah it’s a very important type of data, but it’s,
18:03.45
Denice Ross
But it’s worth thinking about how we manage data at the federal level across all of the types, including administrative data, for example, you know that FEMA collects to give um to to to give benefits to people impacted by disasters.
18:13.11
Jon
Yeah.
18:19.58
Denice Ross
um or the the the the performance data that, for example, the VA collects on its healthcare facilities, that then veterans can use that performance data to find the clinic that’s going to have the services that they need at the quality that they want and, you know, that they can get to by car.
18:26.84
Jon
Yeah.
18:39.32
Jon
Yeah.
18:43.03
Denice Ross
um so there’s ah So I would say like you know thinking about it just as it as a as a statistical unit would is too narrow.
18:50.68
Jon
m yeah yeah
18:54.39
Denice Ross
um And especially because of the the super important role of the geospatial component
19:00.12
Jon
Right.
19:00.79
Denice Ross
and um and and how we need to really start to start to integrate the geospatial um the geospatial component into all of our data.
19:10.07
Jon
m
19:11.77
Denice Ross
and
19:11.80
Jon
All of it.
19:13.08
Denice Ross
Yeah, so um so I don’t i don’t know. Now’s a good time for us as data policy stakeholders to start thinking about what the future should be.
19:19.74
Jon
Yeah.
19:22.30
Denice Ross
And I’m also reflecting, you know, it wasn’t like I’d been in this field for 15 years until I realized I was a data policy person.
19:29.85
Jon
Right. Yeah.
19:31.86
Denice Ross
Right? I mean, like, there’s no, you know, there’s no sort of field of data policy.
19:36.18
Jon
No.
19:36.29
Denice Ross
Right.
19:36.50
Jon
Right.
19:37.14
Denice Ross
Right? where But i mean we know each other when we find each other.
19:37.56
Jon
Right.
19:40.18
Denice Ross
um and some data users sort of you know make the jump to also being data policy and people and and data advocates.
19:40.41
Jon
Yeah.
19:49.78
Denice Ross
and i And I think that’s something that more data users are going to need to do. We can’t take the data flowing or being high quality or relevant to our current needs. We can’t take that for granted anymore.
20:01.35
Denice Ross
And so anyone who’s a data user needs to also be a data policy person.
20:01.65
Jon
Mm-hmm.
20:05.24
Jon
Yeah. I’m definitely looking forward to you having like the endowed chair, the Denise Ross endowed chair at the university of whatever for like the data policy department.
20:15.37
Jon
Right. Like, right.
20:15.96
Denice Ross
I love that. Yeah.
20:17.11
Jon
Yeah. i mean, that sounds like a good, a really good degree. You’re like a data scientist, but you’re doing public policy. You’re figuring out the best technologies to use. I mean, that, that, that sounds like a pretty, pretty good degree to me.
20:29.48
Jon
So yeah, so um ah I wanted to ask, and we’ve talked about this already, but but a lot of things have happened um in the in the in the new Trump administration. They’ve taken a lot of data sets down. They’ve announced stopping a lot of data of different surveys. They’ve put some data back up, and some people are sort of trying to figure out, like, have there been changes to those data? Yeah.
20:55.00
Jon
But I wanted to ask, and you’ve already touched on this a little bit, but but maybe like for people who are not data policy wonks, um what are the implications of the loss of these data, these data tools, um and even just the websites that you can’t get to anymore?
21:11.35
Denice Ross
Yeah.
21:13.82
Denice Ross
Yeah. it’s um It’s worth stepping back a little bit and ah defining what we mean by data because the the scale of the losses so far are different depending on whether you’re talking about a primary collection of data
21:33.02
Jon
Mm-hmm.
21:33.69
Denice Ross
versus a data product like the National Climate Assessment or the Environmental Justice Screening Tool or the climate.gov website that had a bunch of different content on it.
21:39.92
Jon
Mm-hmm.
21:45.94
Jon
Mm-hmm.
21:48.26
Denice Ross
And… um And with any administration change, you see the the websites change their focus. um You don’t see like the type of large-scale removal um that you know that we saw with this administration.
22:03.54
Denice Ross
But you you always see changes in the interpretation of data across administration changes and and and the and and also um you know what data are focused on depending on what the administration’s priorities are.
22:08.39
Jon
Mm hmm. Right.
22:18.97
Denice Ross
um And so that level of churn we would expect to see. um
22:22.67
Jon
Yep.
22:24.06
Denice Ross
Data tools, interestingly, um some of them, for example, FEMA’s Future Risk Index, was that was launched just in December of 2024 and then was terminated like two or three months later.
22:41.70
Jon
Mm
22:41.86
Denice Ross
that and that
22:46.68
Denice Ross
That tool was a really important shift for FEMA thinking about not just risk based on historical events, but risks based on the best science about what the future likely risks are.
22:57.56
Jon
hmm.
23:01.69
Jon
Right.
23:02.65
Denice Ross
um and And that tool being terminated certainly is ah puts more burden on the private sector to to take their best guesses.
23:14.59
Jon
Right.
23:16.25
Denice Ross
um
23:16.38
Jon
Yeah. I mean, yeah. Go ahead.
23:19.13
Denice Ross
Oh, but I was just going to say, and then the, um and then, and then so so there’s, but there’s, so there’s been a lot of removal of those, those data products, the derivative works. um
23:27.83
Jon
Yeah.
23:29.02
Denice Ross
Some would be expected. Most of it is, is sort of above and beyond what we normally see with transitions. What’s been on unnerving for people like me is when those primary data data collections start to disappear.
23:45.62
Denice Ross
um Now, the when we look at the the primary data that have disappeared, though the number that have been actually explicitly terminated is really quite small at numbers in the dozens.
23:57.53
Jon
Mm-hmm.
23:58.26
Denice Ross
If you look at the number of web pages that were taken down, it numbers in the thousands.
24:02.52
Jon
Right.
24:03.26
Denice Ross
um and ah and so And where we are right now is it’s it’s really a ground game, data set by data set. ah And there are different risks that did the data sets experience.
24:17.21
Denice Ross
Some data sets, like the nation’s report card, the National Assessment of Educational Progress, are statutorily required.
24:20.58
Jon
Mm-hmm.
24:25.30
Denice Ross
So they might be, they’re more protected than a data set like the police use of force collection, which is not statutorily required.
24:31.95
Jon
Right.
24:36.06
Jon
But could Congress, uh, presumably Congress, right. Could pass a law that would sexually require any of these data sets. Right.
24:47.96
Denice Ross
Correct.
24:48.82
Jon
Yeah.
24:48.95
Denice Ross
Yeah. And that that’s why I think, you know, sort of a whole of government approach and a whole of society approach makes sense moving forward.
24:52.54
Jon
Yeah.
24:55.35
Denice Ross
And we certainly don’t want to overprescribe a methodology, right, it in um in in law.
25:02.07
Jon
Right.
25:05.27
Denice Ross
That’s not useful, but it is it is useful to. So, for example, the the National Assessment of Educational Progress specifically mentions the demographic attributes that need to be collected.
25:18.36
Jon
isn
25:19.16
Denice Ross
and And that’s useful and meaningful.
25:19.48
Jon
Right.
25:21.68
Denice Ross
And i believe that that collection instrument over time has added additional instruments. So with any given data set, you might have a patchwork of some some pieces of the data set might be statutorily required and other parts might not be.
25:36.86
Denice Ross
And then and then we’re also starting to see some examples um ah where the data where agencies are starting to add questions.
25:47.96
Denice Ross
For example, the one at the post-secondary survey recently.
25:53.78
Jon
it For people who… I’m sure there are a lot of people who take a lot of these data for granted. and So, oh, I can just go grab tabs of education or ACS or whatever it is.
26:06.01
Jon
but But, and so their reaction to this whole conversation may be, well, the private sector has a lot of data. um We saw during the government shutdown of the fall that like ADP and Indeed and a few others were putting out like job market numbers.
26:20.34
Jon
What is, why, or should Well, I’ll put it this way. Should people be nervous about a world in which private sector data are expected or preferred or people want them to replace the the federal, the publicly funded, the publicly created data?
26:40.92
Denice Ross
Yes. ah any Any of the private sector companies that were filling in the gap during the shutdown, ah all of them said that this is not a replacement for the official federal statistics.
26:52.79
Jon
Yeah. Okay.
26:53.32
Denice Ross
I mean, just to be clear, like they they were very explicit about that.
26:53.71
Jon
Okay. So, right. Yeah. Right. Yeah. Yeah. Yeah.
26:57.06
Denice Ross
um
26:57.14
Jon
yeah
26:58.01
Denice Ross
And if you think about it, the the the the the longer you go with that without that official benchmark, the further from reality you can veer. And especially with what’s unique about the federal approach, right, is that it it has to cover all ah all of America, all places and all people.
27:08.38
Jon
e
27:16.38
Jon
yeah
27:17.75
Denice Ross
and not leave anybody behind. But private sector entities have different forces that act on them. I recall after Hurricane Katrina, we were using Google Street View data to see if we might be able to assess housing condition after the storm, like sort of, you know, across time.
27:37.03
Denice Ross
and it quickly became apparent that the Garden District and the French Quarter had more current imagery than like Vallage de Leste, where the Vietnamese community lived in New Orleans East.
27:51.19
Denice Ross
And so there was this data discrepancy ah for different places and different populations. And when you have data discrepancies, it just amplifies the structural disparities that are already in place.
27:58.20
Jon
Yeah.
28:05.40
Jon
Mm-hmm.
28:06.52
Denice Ross
and um And so that’s that’s why it’s it’s so important that we we continue the you know the official like federal data collections. um But we are you know we are going to need to supplement with private sector data and incorporate those into the different data products.
28:26.23
Denice Ross
um And there are, but ah but I also, there’s some exceptions here too, right? Like um I remember after Superstorm Sandy, I was in Department of Energy as a Presidential Innovation Fellow.
28:37.14
Denice Ross
And um President Obama was very frustrated that he didn’t have visibility on which gas stations were open and able to serve people. You might recall the um the footage of like angry people, lots of cars, you know, trying to get gas and um A group of high school students like crowdsourced which gas stations were open and made a map.
28:46.59
Jon
Yeah.
28:51.64
Jon
Yeah.
28:57.66
Jon
Yeah.
28:57.99
Denice Ross
You know, it was it was the height of the civic tech and a movement. And um and Energy Information Administration at the time was literally calling gas stations to see if they were open and if they had fuel.
29:10.81
Denice Ross
um And you can imagine, like, you know, here’s this gas station owner who’s under siege. He’s got all these angry customers out front and the federal government’s calling him to ask if he’s open.
29:18.04
Jon
Just call
29:20.29
Denice Ross
um And so, like, a much better solution is the data that’s already collected and and is very robust through the private company called GasBuddy.
29:32.86
Denice Ross
It’s, you know, and GasBuddy like has this crowdsourcing and it’s used heavily by people who drive for a living, like, um you know, Uber and Lyft drivers and truckers.
29:41.27
Jon
Right.
29:41.48
Denice Ross
And um and when they stop and get fuel, they they they they submit the price of gas, the price that they paid. And then it incentivizes them.
29:51.60
Denice Ross
Then they get you know coupons to buy something in the in the little store at the gas station.
29:56.00
Jon
Right, right.
29:56.40
Denice Ross
And so during blue skies, that data is really high quality. And there’s already a very strong user base. And so when a disaster happens, that also is probably the best source of information for open gas stations.
30:12.28
Denice Ross
Now, that you know that I’m sure the National Security Council doesn’t like having to you know cite GasBuddy as their source ah for gas agencies.
30:16.86
Jon
Yeah. Yeah, probably not. Yeah.
30:22.04
Denice Ross
And it certainly like makes us as a as a as a public a little bit vulnerable if um GasBuddy decides not to share the information.
30:30.74
Jon
Yeah.
30:31.35
Denice Ross
And so that’s like that’s an interesting next frontier, right? It’s like, how do we how do we identify those what our information needs are as a public? And What is the federal government best at collecting and what are other entities best at collecting and how can create we create a national data infrastructure that meets our needs for the future?
30:53.69
Jon
Yeah. I just want to add to that. The combination of those are not, or these different data are not just valuable for researchers, right? Like if you are a person who wants to open a gas station, being able to use GasBuddy to see where the gas stations are open and also to use Census Bureau data to see where people live and the median income and how long people have to commute to work, like the combination of those two data sets would help you make a decision about where to open your next gas station or something like that, right?
31:20.96
Denice Ross
Mm-hmm. Yep. Yep.
31:29.22
Jon
And then for and and for local transit companies where they should put a a bus line or something like that. So um I often think that, that we all get stuck in our little, our little Avenue or whatever, and forget that, you know, combining data, or there are just different needs and different actors in the, in the economy. Um,
31:50.97
Jon
I wanted to ask, ah you mentioned the the civic tech and the crowdsourcing. I wanted to ask you specifically about what’s happening right now. um And you already mentioned the the two sites that that you work on, the data index and America’s Central Data. um How are other people, organizations stepping in to do the similar kind of work that you’re doing?
32:13.34
Denice Ross
Yes. The, um, So we’re we’re monitoring and telling the story about why the why the data matter. And others, especially those who work in specific domains, like Climate Central, for example, when NOAA announced the termination of the billion-dollar climate and weather disaster data set, Climate Central hired the NOAA researcher behind that.
32:38.55
Denice Ross
and And that…
32:38.71
Jon
Yeah.
32:40.79
Denice Ross
we we we we We call that a data set, but it’s actually sort of a derivative product of a bunch of different, mostly federal data sets.
32:45.66
Jon
yeah
32:49.34
Denice Ross
And so as long as those those those inputs keep flowing from the federal government, It may be that Climate Central is actually a better home for the billion-dollar disaster data set. they um you know you know they’re They’re really good at sort of that last mile of making data actionable and you know and and locally relevant. They’re looking at adding features to the data set, you know maybe maybe making it ah the million dollar disaster data set.
33:15.11
Denice Ross
So it covers more and more disasters. Right.
33:17.40
Jon
yeah
33:17.43
Denice Ross
um So that, you know, maybe, maybe that’s, it’s, it’s better home or maybe in the next administration, you know, NOAA decides to, to sort of take up the reins again.
33:28.57
Denice Ross
um That is in contrast to, for example, what Partnership for Public Service is doing around the Federal Employee Viewpoint Survey.
33:37.29
Jon
Yeah.
33:37.40
Denice Ross
ah that That’s an incredibly important survey of of federal employees that happens every year. across the agencies. And this has been the most distributed disruptive year, I think we could agree, for the federal um for federal workers.
33:51.24
Jon
Yeah.
33:53.22
Denice Ross
And if we ever needed to understand, like, how are they feeling? Like, what’s going on?
33:58.65
Jon
yeah
33:58.74
Denice Ross
It’s this year. um So it was really a shame that it was it was canceled for 2025.
33:59.93
Jon
yeah
34:03.74
Denice Ross
Partnership for Public Service saw that gap. and stood up their own lighter weight version of the survey, but they don’t have the email addresses for all the federal employees, right?
34:15.07
Jon
Right.
34:15.12
Denice Ross
So they just don’t, they like, they literally are not positioned to do a comprehensive survey and they can’t replace the Office of Personnel Management’s role with FEB.
34:20.15
Jon
Yeah.
34:25.43
Denice Ross
So it’s ah it’s ah it’s an important stopgap measure, but um but not at all a replacement for the federal data.
34:33.21
Jon
Yeah.
34:35.42
Jon
Yeah. um Yeah. There’s just a lot to do and a lot to sort of organize and in our brains. So um before we wrap up, like what, what does 2026 hold for you? Like, where are you focusing your, your energy in this, in this space or other spaces?
34:55.90
Jon
Maybe you’re just going to set up that endowed chair somewhere.
34:58.23
Denice Ross
and Exactly.
34:58.44
Jon
Yeah.
34:59.51
Denice Ross
Definitely building the field is important. we’ve the The bright spot of all of this is that people are not taking federal data for granted anymore.
35:01.30
Jon
Yeah.
35:07.70
Denice Ross
is So it’s a good chance to talk about what we need from a national data infrastructure and how can we build more resilience into it. there’s um as i As I mentioned before, there’s so many different types of data. So i’m um um I just wrapped up a a fellowship with with Berkeley And ah we in the next couple of weeks, I’m going to be launching a field guide to federal data, which I’m really excited about because I feel like the jobs and the weather data just get so much attention.
35:32.27
Jon
Nice.
35:39.61
Denice Ross
um
35:40.02
Jon
Yes.
35:40.41
Denice Ross
And, ah and but but data touch every every pieces of our lives. When I take, you know, when I go camping with the Boy Scouts and you’re going into the forest and there’s that sign that tells you about what the fire hazard is, that comes from a federal data set, right?
35:55.85
Jon
Yeah. Yeah.
36:00.89
Denice Ross
um And it yeah, so just, ah so, So being more explicit about all the different types of federal data so that as we’re talking about what the future is, where we have um we’re not just over-indexing on the statistical agencies, for example.
36:20.33
Jon
Right. um Well, I’ll look out for the field guide because I think that could be a big value to folks. um I think you’ve got your work cut out for you in the next couple of years. So it’d be my, be my guess.
36:34.58
Denice Ross
ah Well, it’s ah it is a team sport.
36:37.53
Jon
ah okay All right.
36:37.82
Denice Ross
ah And i do i think there’s I think there’s a lot of room for um for people to just continue to build value on the federal data that are flowing. And again, like give credit where it’s due.
36:47.83
Jon
Yeah.
36:50.28
Denice Ross
let’s just call Let’s name the data sets that we count on every day.
36:55.57
Jon
Yeah. So last thing, Denise, if people want to participate, help out, team up with you or the other folks that you um work with, you already mentioned the the newsletter. um Where else can they connect with you or or other folks in this space? Like what’s the what’s their best first move?
37:15.16
Denice Ross
Yeah, so sign up for the newsletter. um when we um We often host webinars when there’s ah ah a flagship data set that’s at risk and there’s a policy window where public input could save it. So you’ll you’ll learn about those opportunities to submit federal register comments.
37:32.63
Denice Ross
Also, one fun thing that we’ve been doing is hosting workshops to create those one sentence use cases about how a specific federal data set benefits everyday Americans.
37:46.15
Denice Ross
And so if you’ve got a group of data users and you want to sort of shift the focus from how they use the data to how the data benefit everyday people, those workshops are there an hour long, they’re really fun to do, very satisfying, and then we’ll do, you know, social media campaign that that’s also gets gets a lot of attention.
38:04.88
Denice Ross
um So please reach out through EssentialData.us if that’s something you might be interested in. Or if you just have one story about your favorite data set that you want us to put on the website, drop us a line.
38:13.24
Jon
yeah
38:16.15
Jon
Awesome. Thanks, Denise. Appreciate it Always good to see you.
38:18.49
Denice Ross
Yeah.
38:19.08
Jon
And I’m sure we’ll talk this year.
38:19.19
Denice Ross
Thank you, John. Sounds good.