By the middle of February, epidemiologists already suspected that the coronavirus was spreading quickly throughout the United States. But testing for the virus was so limited that no one knew where it might show up or how many people had already been infected.

On the latest episode of Social Distance, The Atlantic’s James Hamblin and Katherine Wells talk with staff writer Alexis C. Madrigal about why the U.S. is still so far behind much of the world on testing, which should be one of the first lines of defense during a pandemic.

Listen to the episode here:

Subscribe to Social Distance on Apple Podcasts, Spotify, or another podcast platform to receive new episodes as soon as they’re published.

An edited and condensed transcript of the conversation:

Alexis C. Madrigal: Do you want me to just jump in and tell you how this all got going? It’s a pretty crazy story, actually. So Rob Meyer, another Atlantic staff writer, calls me Wednesday night, the first week of March. And he goes: “Imagine we’re on the Army Corps of Engineers beat. And it’s five days before Katrina. Like, what the blank are we doing?”

And he explained to me, “Listen, we don’t have this number [the total count of people who had been tested for the coronavirus in the U.S.].” We thought, What if we just went state by state and just emailed every press officer, scraped all the data off their websites that we could pull, put it into a spreadsheet, and just hit “equal sum,” and got a number for the number of people who’d been tested?

And so that night we split up the states. I took the bottom half of the alphabet; he took the top. And we were able to come up with the first real number. And it was less than 2,000 people had been tested as of March 6 …

We hit Publish on this story. I go to check my email, and a name I hadn’t seen for a long time pops in: Jeff Hammerbacher. He and I were freshmen at Harvard together.

The subject line is “Did y’all use my spreadsheet!? Thanks for writing this up. I’m curious if you used my spreadsheet to help track this data.” So I click on this thing and I’m like, Oh my God, he built the same spreadsheet.

Katherine Wells: Is this publicly accessible?

Madrigal: Yes, it’s publicly accessible on COVIDtracking.com. And the Twitter feed is @COVID19tracking. We started it on Saturday. And it has, like, 90,000 followers now.

Wells: Ninety thousand and one.

Madrigal: I think this is really important—right now we don’t actually know how much testing is being done in the United States. Still. But we know that testing is the key strategy in trying to move back to normalcy. We need to be able to test and test and test. And we needed to be able to track our success …

Wells: So this was a week ago? I’ve lost all sense of time.

Madrigal: This was March 7.

Wells: Two weeks ago. It’s March 20—who knew?

Madrigal: So we put out a call for volunteers, because this is a grueling data-entry process. You go to every single state’s website, and you enter the numbers, and the way we do it is you actually then have a double-checker that comes through and makes sure all the numbers are right.

Wells: So just to be clear, COVIDtracking.com—which is a website created by you, Alexis, a journalist, and your buddy from college—is currently the most comprehensive source of data on testing rates, something that in other countries is being provided by their health departments in the government.

Madrigal: Exactly. It’s a bunch of volunteers working all over the country all hours of the night. And we’re doing what other countries’ national public-health services are doing.

James Hamblin: We have strong, excellent [state] public-health departments, which are helping and being transparent and working.

Madrigal: Oh, 100 percent.

Hamblin: And that’s the way the system is designed to function. So I think that there’s definitely weirdness happening. But it’s partly the design, right, Alexis?

Madrigal: Yeah, I think that’s true. We have a system that’s designed around state-level control of a lot of these things, because we’re the United States. We know that the public-health people out there want to do this. We want to help them.

I feel like what this project has become is an accountability project, where every volunteer hour logged, every cell that we fill out, every time we publish, it’s just a reminder that we’re not getting the information we need and it’s not actually really the fault of our public-health infrastructure. It’s the larger forces around them. And if we can align those larger forces with their needs, then we should do that.

Wells: Is there anything that this data is telling us right now? Or is it not complete enough yet to be able to draw any conclusions about how we’re doing as a country?

Madrigal: Testing is scaling up. We’re testing more people. We’re not sure how much we’re scaling up, because we’re still missing a lot of data. And of course, there are a lot of cases. A lot, a lot of cases.

Hamblin: Alexis, have people talked at all about an antibody test? … What will be really informative is to have a test that can actually tell us how many people had this in the last three months, and just couldn’t get tested or didn’t get tested.

Madrigal: That’s such a good point. I’m so glad you brought that up. If you want to see the battle plan, I refer you to a Twitter thread. I’m sorry, but it’s the best plan that I’ve seen.

Trevor Bedford is a scientist who has been my guiding light through this whole ordeal. Just yesterday or the day before, he basically laid out, Okay, this is how we get back to normalcy. And the answer is basically spinning up this kind of testing as fast we can, getting the antibody tests rolling. And then as soon as we start to get those antibody tests rolling, getting people who have immunity back to restart the economy.

Wells: It’s good to remind ourselves—there is a lot of inspiring stuff happening right now.

Madrigal: I don’t want to overestimate the impact that we’re having. We’re just, like, this tiny little ship that’s rowing along with a bunch of volunteers. But I do think we’re having an impact, at least at some level. I just know that we’re pushing as hard as we can and hoping that we hit the right lever accidentally and we make this situation better.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.