Stories That Matter: How ProPublica's Cezary Podkul Shed Light on Massive Unemployment Claims Fraud Nationwide

While out-of-work Americans battled outdated unemployment insurance systems to file legitimate claims during the pandemic, scammers aided by automated tech tools were cashing in.

Almost immediately after the COVID-19 pandemic began, labor department officials in states nationwide noticed an alarming increase in fraudulent unemployment claims involving identity theft — a trend that really skyrocketed once federal and state programs for enhanced unemployment benefits started going into effect.

In June, Maryland officials said they had identified more than 500,000 potentially fraudulent claims in just the previous six-week period alone, and roughly 1.3 million since the pandemic began. Each of these phony claims represents a potential victim of identity theft, whose Social Security number and other private information was used by a scammer to attempt to commit fraud — often successfully. Even top government officials and lawmakers aren’t immune — fraudulent claims were opened by scammers in the name of Ohio Governor Mike DeWine and U.S. Senator Dianne Feinstein.

In “How Unemployment Insurance Fraud Exploded During the Pandemic,” ProPublica reporter Cezary Podkul reveals the conditions that allowed scammers to commit fraud by filing huge volumes of bogus unemployment insurance claims and delves into the marketplaces operating online in plain site where criminals would buy and sell victims’ personal information. It was, Podkul says, possibly one of the biggest fraud waves in U.S. history: “Depending on who you ask, it was tens of billions and perhaps hundreds of billions, paid out in improper payments, including due to fraud.”

We talked to Podkul, who specializes in data-driven stories, about his process for compiling and analyzing large volumes of information — and got his tips for other reporters who may be contemplating similar projects.

The interview has been edited for concision and clarity.

Seems like I’m seeing reports in my local media about fraudulent unemployment claims several times a week at this point. Is that how this got on your radar?

Podkul: It’s one of those issues where everyone seems to know someone who either had their identity stolen and had a fake unemployment insurance claim in their name or knows someone who has a friend who did.

I had just finished another project for ProPublica in April and then my editor mentioned this as a possibility and asked if I’d be interested in looking into this further. It just seemed really interesting because I’ve been seeing all of these reports of pandemic-related fraud, like with the Paycheck Protection Program. There had already been local stories written about it, but it seemed like the time was right to do a bigger piece that would dig deeper and give people the big picture of everything that has happened with this issue over the past year.

How long did the research and reporting process take?

Podkul: That conversation happened in spring, probably around May, and then all in, it took about two months to do this story — which was pretty quick. The way it all came together is, I like to follow a multi-thread process. So at the same time I was filing records requests I was interviewing and looking for potential victims to profile and trying to gather as much data as possible and collect it all at the same time while doing analysis and setting up additional interviews. My general approach is to try and have everything progressing simultaneously.

One thing that helps is the strength of ProPublica’s pitching, vetting, and editing process, which is very efficient, and helps because you get feedback all along every step of the way so you can see what may be missing, and that enables you to be very efficient in your work.

I imagine you had a lot of public records requests.

Podkul: For this story, there actually weren’t that many because I realized early on in my reporting that unemployment insurance claims are not public records. They are considered private information because they involve things like Social Security numbers, so public records acts don’t apply. So I knew right away that it would be a waste of time to try and fight that fight. I did file some public records requests, but I was very strategic about it, so instead of making broad blanket requests, I focused on asking specific questions. For example, with the Office of the Inspector General in Washington, D.C., I had learned from my reporting that the number of full-time investigators they had on staff had been decreasing for many years and I wanted data that would show that timeline. And that data was covered under public records.

One of the things you have to be mindful of is the workload for staff at these agencies. Given the national emergency in which they were operating, where they’ve got backlogs of claims and all these people still waiting to be paid, and then they’re dealing with this huge wave of fraud, this didn’t feel like the time to be going on a big FOIA fishing expedition.

Outside of that, it was just a lot of — I hesitate to use the term “shoe leather reporting” because I can’t remember the last time I actually did shoe leather reporting — but it was a lot of working the phones and interviewing. And then once I learned where the action was, jumping on Telegram.

Did you find that state officials were hesitant to talk to you, afraid that shedding light on this issue might reveal, in some cases, how easy it seemed to be to commit fraud?

Podkul: That was certainly a concern, but to their credit, for most of the states that I contacted, they engaged and provided responses in a timely manner. All of them made an effort to answer as much as they could. Whenever I deal with a situation like this where you’ve got 50 states each running their own unemployment systems and the territories as well, you’re talking about 53 systems to implement one national unemployment program. In a situation like that, my first stop was the National Association of State Workforce Agencies, and I was a little surprised that I never got a response from them. But instead it was actually the states that engaged, which was a little unexpected. I’m glad the states did engage, though, because I think it’s very important in a story like this to hear what they have to say and how they’re dealing with it.

Once you gathered all of this data, how did you go about analyzing it and focusing on the parts you wanted to include?

Podkul: I’m very fortunate at ProPublica to get to work with [editor at large] Allan Sloan on a bunch of projects. Allan and I just have a really good working relationship. We developed this process a while back where, as we pick up interesting things that we think are part of the story, we just start writing a memo. I used that for this story, writing a very lengthy memo to myself — I think it ended up being like 100 pages — with all of my notes and the interesting tidbits from these reports. A lot of the reporting involved reading government reports from the Inspectors General, reports from other sources like the Century Foundation, the Government Accountability Office, and putting all of that into on document and organizing it into sections — listing data under the points it supports — and going bullet point by bullet point, and takeaway by takeaway.

Once you’re done with that, you have this long document, so you go for a walk and clear your head, and then you start to really read it until you start to see the big picture. Once you put everything in one place, it enables you to see the big picture more clearly.

A cruel ironic twist to this is that we were hearing so many reports of people who were trying to file legitimate claims and were having trouble — particularly in states like Florida with notoriously outdated or inefficient unemployment claims systems — and yet it seemed to be so easy for scammers to file all of these fraudulent claims. Were you surprised at how easy it seemed to be for them?

Podkul: That surprised me at the outset, just how easy it was. And just to be clear, as time went on and states caught on to the fraud, they did try to lock things down and implement tougher safeguards, but certainly at the outset it seemed like the door was wide open and billions of dollars in these fake claims got paid out. Estimates vary as to how much was actually paid out but no matter who you ask, the answer is too much. Especially in a situation like this where so many people are still suffering and waiting to get paid on legitimate claims. So yeah, it was absolutely too easy for criminals to get through and have their fake unemployment claims paid out. On the flip side, the question is, is it too hard for legitimate claims to go through the process? If people with legitimate claims are waiting and waiting while criminals are getting through, something in the process is broken.

Did your reporting identify anything people can do to protect themselves from being the victim of a fraudulent claim filed using their identity?

Podkul: There are practical things people can do. The raw fuel behind this wave of cyber-crime was stolen identities and they have to come from somewhere. There are various ways that criminals steal people’s information, which they then sell and resell in various forums. One of the big ways is phishing, so it goes back to the old advice of not clicking on an email that looks suspicious and not opening random attachments. I just did another story on fake job ad scams and that’s another way that criminals are getting people’s identities, which is particularly nefarious right now because so many people are out there looking for jobs.

Be smart about what you share online. Use a password protector, don’t use the same passwords everywhere, and be careful about providing too much information online, because so many of these websites end up getting hit with a breach of some sort and getting their data stolen.

What do you think it is about Telegram that seemed to make it such an attractive platform for these scammers?

Podkul: It’s probably no different than with a lot of messaging apps, in that people want anonymity and the ability to securely exchange these types of messages. It’s anonymous, encrypted, and gives people this ability to host channels where they can attract lots of users, gain a following, and blast messages out to lots of people. And then you can also automate things. You can create bots to send automatic responses to inquiries. So there’s a lot of functionality, but I think part of it is also that it’s fun for them. There are stickers and emojis and all kinds of ways to attract attention.

All of those Telegram conversations that we mention in the story, we did bring to Telegram’s attention and we never heard back. We asked for a comment and got no response. But I did notice that, after we started asking those questions, a bunch of the channels where people were exchanging those kinds of messages were taken down, so it does look like the company took some action. But there are still lots of them out there.

Any tips for other reporters working on similar types of stories?

Podkul: One piece of advice I have for journalists is, don’t be afraid to cover a subject or go deep on a story just because it’s been covered elsewhere to some degree. With unemployment insurance fraud, there had been stories written about the issue. It was popping up in local media, warning people to be on the lookout. There’s a lot of good work still being done by local TV stations and newspapers, even with diminished staff. So if I see good work being done on a topic, I don’t see that as a reason we shouldn’t do a story. If anything, that’s all the more reason to devote some resources and try to elevate the issue.

One of the first things I do when I’m thinking about doing a story is do a clip search and see what’s already been done. And if I see a string of local stories but don’t see the big, broad national story, that’s a great sign that there’s room to do that.

On the data side, whenever you’re dealing with 50 states, it’s helpful to try and find a central repository where you can find all the data. In this case, the Department of Labor Employment and Training Administration is where all the states report their data on unemployment claims, so that was a great repository, and the Century Foundation had put together a data dashboard with a lot of that data as well. Try to find centralized sources of data that you can use instead of having to go state by state. That really helped us speed up the analysis.

Stories That Matter is a series of interviews with the people behind some of the best and most influential journalism being done today, focused on reporting, writing, and lessons we can learn from the process of creating great work.

---

The Postscript

Additional content and context, added to everything we do.

Predict: What Happens Next?

“At some point whenever the pandemic is over and there’s a postmortem analysis of how this process went, some of the questions will go toward what went wrong and how not to repeat those mistakes in the future,” Cezary Podkul says. “And those are conversations already happening at the Department of Labor as they think through all of this stuff.”

Read: More of ProPublica’s Pandemic Investigations

• Internal Revenue Service records show that dozens of super-rich Americans received aid intended to help the poor and middle-class: “These Billionaires Received Taxpayer-Funded Stimulus Checks During the Pandemic,” November 3, 2021

• Cheap COVID tests that people could take at home for quick results are in high demand, but many Americans are finding them tough to afford — if they can find any: “Here’s Why Rapid COVID Tests Are So Expensive and Hard to Find,” November 4, 2021

• During the first year of the pandemic, teenagers across the U.S. struggled to complete schoolwork and cope with the stress of a pandemic — but their experience varied widely depending on where they lived: “The Lost Year: What the Pandemic Cost Teenagers,” March 8, 2021

Meet: About the Author

Bobbi Dempsey is a freelance writer whose work has appeared in the New York Times, the Washington Post, Harper’s, and other outlets. She can be found on Twitter at @bobbidempsey.