There’s no doubt that state unemployment insurance (UI) technology needs help. While that’s been true for years, it became evident to anyone paying attention a year ago when people suddenly unemployed by COVID-19 found themselves taking days and weeks and months to apply to and receive UI benefits. Some are still waiting. That was unconscionable, and it needs to be fixed.
The federal government is offering what it says is a solution to this problem, the Office of Unemployment Insurance Modernization, announced on Aug. 31. It is intended to implement the federal UI Modernization Plan, which is in turn based on legislation introduced in March by a number of several members of Congress.
But like so many things proposed by the federal government, it has the effect of wresting control away from the states under the guise of efficiency.
The original plan for UI modernization, as the Department of Labor described it this summer, called for developing centralized infrastructure — an open, modular platform that defined each module necessary to run the UI system while still allowing states to choose which modules they wanted to plug in. By defining the interfaces, states could choose different modules from different vendors. That way, the federal government could encourage a vendor ecosystem, based on easily updated modules rather than on a monolithic “UI system” destined for failure.
Moreover, with such a modular platform, private industry could compete to produce the modules, resulting in much greater competition and leading to better features, service and price.
Here’s the problem
But that’s not what’s actually happening. Instead, the Wyden-Horsford Unemployment Insurance Technology Modernization Act — and now the federal UI Modernization Plan based on it — creates a single government team within the U.S. Department of Labor to build one system that all states are supposed to use.
It’s not going to work.
States are too different to just give them a cookie-cutter solution and tell them to adapt it to their individual needs. And too many UI modernization projects have failed, such as Kansas, Virginia, and Florida. Additionally, who is going to adapt the modules? The states? They did not write the modules; they have no idea what needs to be done to modify them, nor the resources to be able to do it even if they did know.
Moreover, the vast majority of the funding currently designated for the UI Modernization Plan in the current version of the reconciliation bill is either provided to the federal government to write it, or to the states to integrate with it. There is little, if any, funding provided to the states to modernize their own systems. Starved of other resources, they will be forced to accept the federally mandated system.
That’s assuming the system even works. Government agencies are notoriously bad at implementing and maintaining systems this complex. And make no mistake, UI systems are extremely complex, even within one state. The track record for UI implementations is horrible: so many failed state implementations and failed consortium implementations, due to following decades-only monolithic software development methods.
Bad results after spending tens of millions of dollars will not be improved by spending hundreds of millions.
Plus, when multiplied over all states, as the UI Modernization Plan is attempting to achieve, it is way outside the bounds of a single agency’s capabilities. And while state UI systems certainly had problems, there’s no guarantee the federal government is going to do any better. Consider the Small Business Administration, the Internal Revenue Service, and Healthcare.gov.
Having one team of people — especially if they’re in the federal government — in charge of operating and maintaining all of the state-based content is a bad idea. States should be allowed to choose their own content creators and providers; they should not be at the mercy of a handful of people on a team in Washington, D.C. State-chosen providers, whether internal or external, should be creating state content. Locking all states in to the same team severely limits state flexibility, vendor diversity or competition, and innovation in the space. This concentrates power for a multibillion-dollar market in a very small group of people. This typically has not ended well in the past.
Finally, it’s clear that the real goal here is taking local control of unemployment insurance rates and policies away from the states and implementing a top-down federal mandate, under the guise of improving the technology.
If COVID-19 has taught us anything, it’s that states don’t like federal government mandates. And throwing more money at it will not fix this problem.
What’s the solution?
It’s not too late. State-based stakeholders who want to maintain a modicum of local control in their UI systems can tell the Department of Labor and their Congressional representatives that this isn’t how they want to operate. Let the federal government define a platform, interface definitions, the modules needed and security. Let states determine which modules they want to use, and where to get them.
Finally, encourage vendor participation. Any vendor that can implement the service interface should be able to provide content that sits on the platform. By promoting this open, modular platform, the federal government can encourage a robust, competitive vendor ecosystem based on content modules rather than on a monolithic “UI system.” Many more vendors will compete to produce smaller modules, based on their expertise and their familiarity with the needs of the individual states. More competition, and a modern, modular development process, will result in improved user experience.
State sovereignty is at stake.
— Gregg Tahmisian is CEO of Solid State Operations, a Boise-based developer of unemployment insurance software.