States should look beyond Direct File to keep the spirit of reform alive

08. 28. 2025 • Chris Given and Gabriel Zucker

Many state tax administrators approached Direct File with a healthy skepticism — could the IRS really deliver in a matter of months, and would it be worth it for states to join on such short notice? By the program’s second year, though, most of them were enthusiastic. State partners found that Direct File returns were exceptionally accurate, and Direct File put tax administrators in a novel position: receiving fan mail. In integrated states, 89-98% of taxpayers reported being “satisfied or very satisfied” with their state tax filing experience. Said one taxpayer, “I will recommend it to everyone I talk to.” Another taxpayer, asked how the experience could be improved, said, “You can’t improve on perfection.”

With the federal government abandoning Direct File for now, many state leaders are, admirably, looking for ways to keep the party going. But as the field works to advance the cause of equitable tax system access, it’s important to remember that our goal isn’t just to keep Direct Filing. Our goal is to keep making the sort of transformative progress that Direct File showed was possible: delighted taxpayers, increased trust in government, efficiencies and cost savings, and improved access to tax credits for those who need them the most. In order to get these results without the IRS, states should look to replicate the mindset that brought Direct File to life, rather than simply looking to replicate Direct File itself. 

Meeting this moment requires imagination, and as with Direct File, the courage to try something new. We believe this is a time for experimentation and innovation, and states can prove out ideas that the IRS could someday adopt and scale. And there are opportunities to improve state-specific processes and experiences, entirely independent of what the future holds for federal tax administration. 

We’ll illustrate this mindset with three examples of experiences that states could create today. These ideas all leverage the unique position of states to craft moments of “wow” for taxpayers, and they could pave the way for Direct File to return better than ever. But these are just a few ideas. An empowered product team working closely with taxpayers, like the teams that designed and built Direct File, could invent many more ideas and far exceed our limited imagination.

It’s your money, even if you didn’t claim it

Prior to Direct File, one of the IRS’s most trust-building experiences was the math error notice, a letter correcting figures on a taxpayer’s filed return. No, really. Sometimes the notice informs taxpayers they owe money — a necessary, if unpleasant, experience. But just as often, the notice says that the IRS believes you are due more money than you claimed, and wouldn’t you know it, they have already sent you that money without you needing to do anything at all.

What better way to demonstrate that the IRS’s mission is the fair and impartial administration of the tax code (not extorting maximum revenue) than to send back money without you asking. Taxpayers who have this experience just once talk about it for years or even decades after the fact. As lobbyists attempted to spread fear, uncertainty, and doubt about Direct File, these ordinary taxpayers chimed in to social media discussions, rebutting industry talking points with their personal story of how the IRS once sent them a check for twelve bucks.

Most states use math error processes at least sometimes, but most could find ways to expand this practice, issuing more taxpayers more seamless dollars that they deserve.

• Automated correction of under-reported withholding and estimated tax. Taxpayers may understate the amount of taxes they have already paid, by losing track of an estimated tax payment, or underreporting the withholding on their W-2. States are usually positioned to catch such errors fairly easily, since these payments go straight to the tax department. And correcting such returns is relatively straightforward, since the additional payments are returned, 1:1, to the taxpayer. Even if such mistakes by taxpayers are rare, they do occur, and they are a huge deal for those taxpayers that make them. If they are not doing so already, states should automatically detect and issue such refunds.
• Automated detection of unclaimed credits. Some states have tax credits on the books whose eligibility can be conclusively determined based on other information on a tax return, like dependents, income, and/or eligibility for federal credits — and taxpayers may fail to claim these credits on their returns. These corrections may be a step more complex than detecting the payment mismatches discussed above, but are still doable in many cases. If they are not doing so already, states should automatically detect such credit eligibility and issue refunds for the unclaimed credits.
• Faster and more automated processes. Some states may already be doing both of the above automations — but perhaps through manual and time-consuming processes. In this case, building more systemic automation, so that these processes do not require costly human intervention, would save costs in the department and speed up processing for taxpayers, thereby improving trust still further.

Processes like this live outside the realm of filing a return, but are nevertheless a critical aspect of the filing experience for taxpayers. Improving the holistic experience of filing is a critical muscle for states to have independent of any reform or lack thereof of filing processes at the federal or state levels.

Year-round support for gig workers

Some of the taxpayers with the most challenging interactions with the tax system are gig economy workers, who not only file complex returns but must proactively make estimated tax payments throughout the year to avoid an unpleasant surprise at tax time. Every state in the nation has at least tens of thousands of these workers, often earning non-employee compensation from multiple companies. The task of estimating your future tax bill is challenging, particularly when your income is highly variable. And unlike your annual return, it’s a task that must be completed four times a year.

States are well-positioned to make this task easier for their gig economy workers. Imagine a state tool that a worker logs into every quarter. They add their latest income information, either manually, or by automatically importing the data from the companies they work with. The tool combines this information with information from the worker’s prior year return to estimate the amount they should expect to owe, and it helps them make state and federal payments. Email or text reminders would ensure that taxpayers don’t miss a deadline.

Any such tool would best be paired with an outreach and education campaign, taking advantage of the state’s official megaphone, perhaps working in partnership with gig economy companies. Taxpayers would benefit from a trusted tool that takes the headache out of an intimidating task and reduces the risk of unexpected financial shocks. States would benefit from increased compliance, fewer debts sent to collections, and, like with Direct File, an opportunity to position themselves in the role of helping taxpayers get it right, instead of telling them when they’ve gotten it wrong.

In the long run, gig economy workers would want their estimated tax compliance problems solved at the federal level as well; and so, in the long run, it may make sense for parts of such a project to be adopted into a centralized national tool. But this is a nearly-wholly unexplored product space. By implementing and iterating on solutions at the state level now, states would move the field forward and vastly accelerate the potential development of a national-scale solution in the future — all while delivering big material benefits to taxpayers now.

Welcoming back lapsed tax filers

Suppose we were to build the tax filing system of our dreams, in which filing a tax return were as easy as possible and absolutely free. Even in this world, some people wouldn’t file. They’d still have life disruptions that get in the way, and some of them might get stuck in a rut, scared to start filing again, for fear of consequences for not filing or assuming they’d already forfeited their refunds. And, in plenty of these cases, the government would still owe them money, in credits or excess withholding.

Make no mistake — in the American tax system, it is not as easy as just issuing checks to these taxpayers. The reason taxpayers need to file is that their refund depends on all kinds of details that the government does not know and has no way to know, at least promptly. But conversely, it’s not like we know nothing about these non-filers. This is especially true for taxpayers’ prior-year returns. Unlike the current tax year, the dust has largely settled on past tax years. States can be relatively confident they have a nearly complete picture of a taxpayers’ income and withholding via information returns, even those that arrive late. States can see whether another taxpayer has claimed a given non-filer already, and if the dependent they claimed in a previous year, or the spouse they filed with in a previous year, has appeared on another return.

Most importantly, once more than a year has gone by, the priority might no longer be to ensure the taxpayer files a full return claiming every single dollar to which they are entitled. If the likeliest alternative outcome is that the taxpayer gets nothing at all, the priority becomes to ensure they at least get some of that money, and to welcome them back into the tax system.

For these taxpayers, who otherwise may get no money at all, states can automatically generate their best guess of the taxpayer’s prior-year return, using information returns, returns from earlier years, and information from other returns filed in the reference year. (If the taxpayer filed a federal return for the reference year, states can use that return too.) States would send these automatic returns to taxpayers, who could simply confirm a few key facts to claim their money. Taxpayers could always choose to file a full traditional return themselves.

The benefits for a non-filer of receiving a pre-filled prior-year return could exceed the monetary value of the refund. If the experience is done right, such a return could be a warm invitation to return to the fold and file in the current year as well, getting the full raft of credits to which you’re entitled. It could relieve you of feelings of guilt and stress over non-filing, and demonstrate that you’re not in trouble and that the state is on your side.

Moreover, by experimenting in the more risk-tolerant environment of prior-year returns, states can prove out the potential for these ideas to be brought to current-year returns. This work not only serves a short-term benefit for an important audience, it provides a stepping stone for a long-term vision of moving U.S. tax administration toward a more return-free-like experience. 

As with the second experience, this would not only provide concrete material benefits in the short term, but break new ground for a tax filing reform to be scaled nationally in years to come.

These are all ideas that can productively advance the tax system without being held back by an IRS that has turned its back on taxpayer service. They are processes that could exist entirely at the state level, where a taxpayer might plausibly use a standalone experience, or they are ideas that would break new ground in tax reform, pushing the envelope of what’s possible and providing models for future federal reforms.

But, again, they are just ideas. States needn’t adopt these specific ideas. Our call to states is to carry forward the mantle of Direct File not by copying it, but by building upon how we did it.