Technology alone can’t fix state unemployment systems

When unprecedented demands caused states’ unemployment programs to fail during the pandemic, it was easy to blame technology.

Yes, outdated systems were a big part of the problem, but many states were able to quickly replace their call centers and use cloud technology to scale up and meet surging demand. And though modernizing systems and moving to the cloud keeps systems from crashing, the harder tasks are the ones that follow: reviewing applications, assigning benefits, and delivering support for millions of job seekers.

As the economy continues to rebound and companies are struggling to attract talent, it’s worth pausing and thinking about how we can do those things better. But if all we do is modernize old systems without addressing fundamental strategies and objectives, we will be missing an important piece of the puzzle.

Administering unemployment insurance is about applying judgment to the facts in each individual case. States must determine who is eligible for which benefits, while guarding against fraud and upholding the law. It’s not easy or straightforward.

Artificial intelligence, applied carefully, holds promise to streamline some of the judgment work and to identify cases suspected of fraud. It does little, however, to assuage the concerns of unemployed people who resort to the phones to find out what’s become of their applications. As a result, case workers who should be using their judgment to resolve claims are instead answering calls to ease anxiety or help with passwords.

What we need is an unemployment insurance system that serves claimants and reserves the UI professionals for the work only they can do. Such a system proactively informs and updates claimants to ensure they always know the precise state of their claim — where it is in the process and what timeline it is on. This empowers claimants, leading to less need to contact call centers for updates. Likewise, the system must clearly, concisely and accurately explain the process to the claimant to eliminate confusion and concern. When information is missing in an application, claimants should be informed — by text or email, with an easy way for claimants to fill in those blanks.

Like the Domino’s Pizza Tracker, this kind of innovation addresses the customer’s problem. We all can accept that pizzas take some time to cook, but it’s really helpful to know when your pizza will show up at your door. If an unemployment insurance recipient can know when to expect an answer about their claim, he or she can start to make plans and answer basic concerns: Can I pay my rent next week? What about paying for childcare while I am looking for work? If you’re applying for unemployment, you’re already having a bad day. Getting you the information you need to regain control is something that technology – cloud computing, secure data transfers, and smart applications of AI and machine learning – can do remarkably well right now.

The most important thing state agencies can do for the jobless is to help them get a well-paying job where they can find success and career growth. Unemployment insurance is a stop-gap — an emergency solution to an immediate problem. The long-term need is not just money to live on, but the means to earn a living.

Research Improving People’s Lives, the organization I lead, developed a solution that applies ML, AI, and cloud technologies from Amazon Web Services to unlock previously siloed government administrative data and turn it into recommendations for reskilling into new careers. The Data for Opportunity in Occupation Reskilling Solution, or DOORS, combines states’ data on wages, unemployment insurance and job training, and converts it into personalized career recommendations, job matching and connections to job training programs, all in one user-friendly interface.

This is the difference between giving a man a fish so he can eat for a day and teaching him to fish so he can eat for a lifetime.

We all know people who were failed by unemployment insurance, people whose already dreadful situations were made worse when they needed help the most. Let’s set the bar higher and strive to overachieve.

A commitment to treating each claimant like our very best customer is the first step. Employing technology to achieve that objective is the second. The third, and hardest, is to help the unemployed get back to work. Technology can help here, as well. We need intelligent tools to help assess present skills and the acuity to learn new ones. We also need matchmaking tools to help guide job seekers to new careers with long-term potential.

Technology can get us there. But to do so, we must prioritize an obsessive commitment to the best possible user experience. Likewise, our work must always strive toward the ultimate objective of helping the unemployed return successfully to the workforce.

Scott Jensen is the chief executive officer of Research Improving People’s Lives, nonprofit that works with governments to help them use data, science and technology to improve policy and people’s lives.