Exit interview: What the Labor Department's unemployment insurance guru thinks about the state of the system

Michele Evermore left her post as the head of the Labor Department’s new Office of Unemployment Insurance Modernization in December after serving nearly two years within the agency. 

Her career spans Capitol Hill and the National Employment Law Project and more before Evermore went to the Labor Department to work as a senior advisor in early 2021. In mid-2021 she moved to the role at the new modernization office.

Now working as a senior fellow at the Century Foundation and as a visiting, non-resident scholar at Heldrich Center at Rutgers University, Evermore talked to FCW about what is next for jobless aid, whether the system is ready for a recession (spoiler: she says that it isn’t) and how to think through equity and fraud in unemployment. 

Editor's note: the following interview has been lightly edited for clarity and length.

What was it like working on unemployment in the wake of the pandemic?

In a word — intense… 2020 felt like we answered a lot of the questions about pandemic-era benefits. But really, we had a lot more to do in 2021. Making sure those programs operated like they were supposed to was quite a challenge. By 2021, I think people had been working very hard for a long time at state agencies and at the federal level, so making sure that we're being ambitious at the same time as making sure that we're aware of our resources was also something to be thinking about.

What were the biggest challenges in trying make progress and get your work done?

I really think it was trying to deal with these outstanding questions of program operation at the same time as trying to ensure that people got the right benefit at the right time. A really good example here is, what are some of the guardrails around ID verification?

Congress allocated $2 billion to the Department of Labor to improve timeliness, equity and fight fraud in the unemployment insurance system. There's an inherent challenge in trying to figure out what to do with a one-time allocation of funds for a systemic problem… How do you set up ways to address the problem so that you're not creating a funding crisis for yourself down the road?

You mentioned systemic challenges in the jobless aid system. What are those headline problems?

[Unemployment] administration has been underfunded for decades… which created a real problem at the start of the pandemic, when all of a sudden initial claims went from 211,000 to 6.6 million in a few weeks, which is far higher than the highest on record of 695,000 in October of 1982.

Also – because there's sort of a race to the bottom in some ways between states on benefit-level duration, access, all of that sort of stuff – Congress really felt the need to make sure enough people had access to unemployment insurance that it would do what it was supposed to do, and that's really cover people who were hurt by these massive widespread shutdowns. That meant creating a new program with new eligibility requirements and standing it up nearly overnight, increasing duration of benefits for regular UI, and adding this extra $600 a week up until the end of July, which became $300. All of that was a challenge.

Then around May 2020, you really started to see these coordinated international crime rings coming after unemployment insurance benefits, and that was a whole other challenge… Some of the [existing] fraud fighting strategies were geared toward a different kind of fraud, and so they had to immediately adopt brand new fraud fighting strategies, and states remarkably responded very quickly.

One thing people talk a lot about in unemployment is outdated tech in state systems. What do people get right and wrong about unemployment and its technology?

People let that word [modernization] do a lot of work. Modernizing is important. Modernizing in the right way is more important. There are systems that are still operating on COBOL that I would say are doing fairly well. Just because they haven't moved all the way to the cloud, doesn't mean they're not doing systemic improvements… A great example is New Jersey. They're not fully on the cloud, but they're modernizing in the right way… They've established an office to be UI modernization at the state level, and they're figuring this out in an agile, smart way. 

Let’s talk about fraud. What should people consider when they think about fraud in the unemployment system?

The biggest takeaway about fraud in unemployment is that unemployment happened to be the target of the moment. But these fraudsters, in large part, got the information they needed to expose unemployment insurance through other breaches, other security hacks. This personal information, this data is out there… But unemployment insurance was a little bit the canary in a coal mine here. This data could be used in any way, right? That makes it sort of a whole of government, whole of society problem. When people talk about it as an unemployment insurance integrity issue, of course it is, but we shouldn't be thinking of it as just an unemployment insurance problem.

That segues perfectly into the executive order on identity and fraud. If you were authoring this, is there anything you definitely would or definitely would not want to see in this?

I have confidence in the people in the White House who are working on this. I think that they're asking a lot of really good questions… You have to make sure that whatever ID verification methodologies are out there, there needs to be more than one verification opportunity for claimants and that claimant experience is a key factor in making any decisions about whether or not an ID verification system is doing what it's supposed to do.

ID verification has a role to play there, but it's very easy for it to be a bottleneck for folks. There are particular concerns around access for people of color, people with disabilities, people with limited English proficiency, so just making sure that you get it right for all the claimants is top of mind for me.

What is the landscape as far as equity and UI goes?

Here's what I'm really excited about. Over the course of my time in the administration, I really think people got a better understanding of what equity looks like… The equity grants really made me excited. The other thing is the level of sophistication that the tiger teams are bringing to the process. It’s really impressive. They're really getting into the weeds on what does equity actually mean? How you make sure, not just that everybody who gets in the front door is treated equally, but that everybody's able to get to the same front door.

Let’s talk about equity in terms of fraud prevention and identity verification in particular, which I know has been a worry for some advocacy groups. 

In 2020, when I testified in the Senate Finance Committee, I actually pointed out that an overfocus on fraud leads and has led to an increase in erroneous denials… Every previous administration going back a few have been very focused on UI improper payment rates, and to the extent that you tip the scales too much in the direction of caring about fraud over everything else … You can prevent fraud by never paying anybody anything, right? But that's not the point of an unemployment insurance system.

I would say on the other hand, it's really, really important for UI systems to be vigilant against fraud because once people can paint the unemployment insurance system as paying a lot of bad actors, there's a lot less support for this key, very important economic protection for people. 

The most important thing to think about when you know, looking at fraud interventions is to make sure that you're inspecting the fraud intervention itself for bias and that you're balancing out fraud detection with making sure people who get swept up in the dragnet and are innocent have a timely mechanism for for redressing their problem.

The Labor Department has been taking a more hands-on approach with states on unemployment. How did that work?

All of that effort was informed by what states asked for. What states were asking for, particularly mid-pandemic, was greater coordination from a central source and more help … It’s part resources. It’s part technical assistance.

One thing I'm really excited about is that the Office of Unemployment Insurance created two new divisions that I really hope are able to continue past the end of the ARPA funding. The first is the central response division, and that's the one that's right now sort of the backbone for the tiger team efforts… The other is the innovation division - they are going to be the technical assistance team hopefully long term to help states with their modernization efforts.

Is the system recession ready?

No. It's really not… Honestly, my biggest concern is the effect that the pandemic had on state systems in terms of their resources and their personnel. A lot of people left state agencies.

I also think what most people don't realize is the pandemic isn't over for state agencies by a longshot. They're still doing a lot of adjudication of pandemic-era questions… I'm very concerned that state UI agencies are going to be challenged whenever there's another recession, unless the recession is kind and waits a good five years or so.

What’s next for you?

Maybe the first question [I want to take on] is disability access to UI. I really also want to nail down, what are the noncontroversial things we really can do? There's definitely a refocus that needs to happen around unemployment insurance. If you read the bipartisan 1980 recommendations and the bipartisan 1996 recommendations, they are great…I'd love to dig through how they were thinking about things back then and how we can bring some of those ideas forward. 

I definitely am going to continue to spend some time thinking about IT modernization and figuring out not just how states can do better, but also how advocates can more effectively be at work with systems and how claimants can inform the process better and things like that.