The Things That Cannot Be Changed

Direct File was an obvious idea. Many people had tried to make it happen. The Direct File team was good, but we weren’t any smarter than those who came before. So what made this time different?

I’ve always found Reinhold Niebuhr’s serenity prayer to be a useful thought technology for life in government.

God, give us grace to accept with serenity the things that cannot be changed, courage to change the things that should be changed, and the wisdom to distinguish the one from the other.

Success as a civil servant requires an ample helping of all three traits, particularly serenity. However, “the things that cannot be changed” is not a static category, and it rewards assiduous attention. The smallest shift can make what was once an impossibility suddenly possible.

Direct File had long been on my dream project list, but when I began working on taxes in January 2021, my strong assumption was that any attempt to make it a reality, like those before, would doubtless fail. However, a little failure from time to time is healthy, and I figured it would at least make for a good story someday.

After years of jogging items up and down on my mental list of the most likely ways Direct File would fail, I didn’t realize how thoroughly I had internalized this assumption of doom until the moment Dixie Warden filed the first Direct File return on February 1, 2024. She had agreed to allow Bridget Roberts, Jen Thomas, and me to observe her progress through the application. Direct File still had some significant known bugs, so my responsibility was to call the whole thing off if Dixie ventured too close to certain tax landmines. As she signed and submitted her return, I realized that I had not let myself believe until that precise moment that we would ever actually be allowed to file someone’s taxes.

Three years earlier, no one could have anticipated that moment would come. Direct File wasn’t on the agenda. There wasn’t and would never be a grand strategy to mastermind events to bring it about. But events, as is their wont, happened anyway. It’s a stretch to call it “wisdom,” but by setting up at the right place at the right time, we were able to perceive that a window of opportunity had cracked open. These are the three events that made this time different.

1. The pandemic

On March 27, 2020, the Coronavirus Aid, Relief, and Economic Security (CARES) Act was signed into law. One of its provisions was the Recovery Rebate Credit (RRC), more commonly known as stimulus checks, advanced in the form of $1,200 economic impact payments to individuals (plus $500 for each qualifying child dependent). RRC had a subtle but significant difference as compared to a previous programs like that of the Economic Stimulus Act of 2008 (ESA). ESA also provided relief payments to individuals, but it crucially required at least $3,000 of income to qualify for the credit. In the CARES Act formulation, RRC could be claimed by anyone with a Social Security number, even if they didn’t have a cent of income.

The IRS launched a massive effort to get checks out to as many people as possible, working the Social Security Administration and the Department of Veterans Affairs to identify recipients of Social Security benefits, Supplemental Security Income recipients, and disabled veterans to whom checks could automatically be sent, despite the recipient not having filed a tax return. I’m still astounded and endlessly impressed that the IRS and partner agencies were able to pull off this feat with workforces still reeling from the early weeks of the pandemic.

For everyone else, you needed to have filed a recent tax return in order for the IRS to know how to get you your money. But a small group of federal employees at the Department of Treasury masterminded an ingenious policy hack and issued Revenue Procedure 2020-28, which created a new option for anyone who wasn’t required to file a tax return, generally because they made less than $12,200 or $24,400 for a married couple. Instead of needing to assemble the documentation required to file a full return, these taxpayers could instead file a simplified return claiming just $1 of income.^[1] The IRS would understand these to be returns for the sole purpose of receiving an economic impact payment and agreed not to challenge their accuracy.

In order to make it possible to file such a return, Intuit, under the auspices of the Free File Alliance, repurposed the existing Free File Fillable Forms product to stand up a non-filer portal (officially, the “Non-Filers: Enter Payment Info Here” tool). Launched on April 10, just two weeks after enactment of the CARES Act, the non-filer portal enabled millions of additional Americans to receive stimulus checks. An Intuit employee would later describe this work as the most rewarding of his career.^[2]

The following year, the American Rescue Plan Act built on the CARES Act model in its expansion of the Child Tax Credit (CTC). The refundable portion of the credit had previously required at least $2,500 of income, but now, in addition to the full credit being refundable, any taxpayer could claim the credit even if they had no taxable income. Hundreds of thousands of kids who had previously been excluded from the benefits of CTC would now see their share of the money.

In order to reach the children of non-filers, the IRS and Treasury dusted off their CARES Act playbook, and on June 14, 2021, they reopened the non-filer portal.^[3] But after a year, the portal was seen in a new light. For those who have not had the pleasure of using Free File Fillable Forms, it’s a deliberately excruciating experience.^[4] The interface imitates paper forms (don’t try it on a phone) and invites the user to manually type numbers into boxes. The non-filer portal used the same building blocks, so the experience is comparable, just with fewer boxes to type in. That might have been fine for something shipped in two weeks, but there had now been ample time to make it even slightly better.

Congress and advocates began to ask reasonable questions. Can we make this work on phones? Can we get this in Spanish? Can we make it not look sketchy as shit? (Okay, that last one might have just been me.) The IRS’s answer, or what was really the Free File Alliance’s answer: no. Congress didn’t love that answer; neither did the advocates. It’s not a great feeling to go through the rigamarole of passing a law, only for its benefits to not reach those who need them the most because the delivery mechanism is hostage to the whims of a third party incentivized to make that experience as bad as possible. Administration officials were startled to be on the receiving end of this ire.

A senior advisor to Treasury and I had anticipated this eventuality a couple months prior, and we had started a conversation with Code for America about the possibility of them creating an alternative to the official non-filer portal. The result, GetCTC.org, began beta testing in August, and provided the White House with a much needed solution to its portal problem. It worked on phones, its plain language was available in English and Spanish, and for bonus points, it looked like a tool you could trust. The White House enthusiastically promoted GetCTC as part of their CTC outreach efforts.

Even as a nonprofit came to their timely aid, however, officials had learned an object lesson about what it meant for the government to not be in control of its own destiny, a lesson that would not soon be forgotten.

2. Intuit leaves Free File

Those alive at the time will always remember where they were when they heard JFK was shot. I will always remember where I was when I heard Intuit was leaving the Free File Alliance.^[5] It was an event that so completely upended what was possible, I was vibrating for a week afterwards.

Free File was created in 2002 and successfully forestalled the development of Direct File by more than twenty years. The Bush Administration had established “E-Government” as a core pillar of its President’s Management Agenda and proposed the creation of “an easy, no-cost option for taxpayers to file their tax return online.” At the time, the average cost of electronic filing via a third party was just $12.50 ($22.41 adjusted for inflation), which the administration noted was quite a bit more than the cost of a 34 cent stamp.^[6]

The tax preparation industry scrambled for an alternative. Intuit organized its competitors to form the Free File Alliance and offered the IRS a deal: the companies would agree to provide free versions of their products to low-income Americans in exchange for the IRS committing to not offer online filing. Heralded as a “public-private partnership,” the arrangement appealed to the White House, despite some hesitation from the Office of Management and Budget and Treasury.

For its part, the IRS was relieved to not have to undertake a risky, high-profile IT project. The memory of 1996’s Cyberfile disaster was still relatively fresh. In late 2022, I attended the retirement party of a long-time IRS employee and chatted with someone who was closely involved in the early E-Government work. He recalled of Free File, “We were given just nine months to get something out the door.” He gave me a wry smile. “Sounds familiar, no?” He wished me luck; it was clear he thought Direct File would need it. (He wasn’t wrong.)

What happened next has been well documented, including actions the IRS took to limit the scope of the Free File program out of concern for the “future revenues and profits of the publicly traded company participants,” actions that had Free File taken on its own behalf, would have constituted illegal collusion and price fixing. Apparently by design, Free File usage dwindled.

In 2005, the IRS terminated its TeleFile program, which let taxpayers file via phone.^[7] Paper returns were now the only remaining way to file without going through a third party. The success of electronic filing relied on the goodwill of industry. Efforts to prevent identity theft and refund fraud hinged on the voluntary participation of companies, and the IRS was often negotiating from a position of weakness.

Then came 2019, a tumultuous year for Free File. The year started promisingly enough, with the near realization of lobbyists’ long-held dream of enshrining Free File into law via the Taxpayer First Act. However, public backlash resulted in this provision being stripped, and over the coming months, reporters at ProPublica released a series of articles highlighting deceptive industry tactics. Investigations were launched, and by the end of the year, Free File was on its heels. In an attempt to deescalate, Free File and the IRS agreed in December to remove the language from their Memorandum of Understanding (MOU) that barred the IRS from offering Direct File.

Both parties understood that this was for show. Little had changed, and the IRS’s hands remained effectively tied. When I arrived at the agency in 2021, Free File was described to me as “the third rail of the IRS.” Venture too close to that third rail, as the Volunteer Income Tax Assistance program found out when its Facilitated Self Assistance model became too popular during the pandemic, and you could expect to be put on blast by industry lawyers during tense phone calls.

I saw Free File less as a third rail and more as a Gordian Knot. The way it prevented Direct File was not by explicitly barring it, nor via awkward meetings, but rather by making it too hard to do anything else. Despite Free File’s underutilization, there were about four million taxpayers who used it in filing season 2021. Even after the prohibition was dropped in 2019, the MOU still contained the following clause:

Should the IRS commit funding to offer Services for free to taxpayers the IRS shall notify [Free File] immediately. If the IRS gives such notice during the tax season (between January 1 and April 15 […]) of any year, [Free File] may, by written notice to IRS, terminate this MOU, effective on April 16 […]. If the IRS gives such notice between April 16 […] and October 15 of any year, then [Free File] may, by written notice to IRS other than during tax season, terminate this Agreement, such termination to be effective no fewer than 30 days after the date of [Free File]'s notice of such termination. If IRS gives such notice between October 15 and December 31, [Free File] may by written notice immediately terminate this Agreement at any time on or before December 31.

The notification clause allowed industry to take their ball and go home the moment the IRS committed a single dollar to Direct File, instantly depriving some number of taxpayers of the method they used to file. But since the IRS had been sitting on the sidelines for literal decades, it would require years of work to catch up and create a service that could address the diverse tax situations of all of those taxpayers to the standards of the IRS. In the short-term, some people would need to find a new way to file, and industry would point the finger at the IRS. The IRS had no room to maneuver out of the predicament in which it had been trapped.

Then on July 15, 2021, Intuit announced that it would no longer participate in Free File, following the lead of H&R Block in 2020, leaving only TaxHawk (aka FreeTaxUSA) and a number of smaller companies to man the fort. The eminently predictable result was that a million fewer people used Free File in 2022. How did this enable Direct File?

In short, Intuit gave up its leverage. In the chronology as it played out, there are four events. (1) Intuit announces it’s leaving Free File. (2) A million fewer people use Free File. (3) The IRS announces it will pilot Direct File. (4) 140,803 people use Direct File in a limited pilot.

Now imagine a different chronology. (1) The IRS announces it will pilot Direct File. (2) In response, Intuit announces it’s leaving Free File. (3) A million fewer people use Free File, and only 140,803 people use Direct File. I would argue that in this world, Direct File would have been dead on arrival. The IRS would have gone through a whole lot of trouble for the net result of fewer people filing their taxes for free. It just didn’t work politically, and I saw no way of untangling that knot.

But then Intuit cleaved the knot for us. And so when the Direct File pilot happened, Free File saw a significant increase in usage, the result of more taxpayer interest in options for filing for free (and admittedly, a whole lot of Free File/Direct File brand confusion). Taxpayers had more options, not less, and more of them were able to file for free, an unambiguous win.

(Now that Direct File lies comatose, Intuit gets a second bite at the apple, and they and the rest of the tax preparation industry will attempt to restore their leverage over the IRS and state revenue agencies, lobbying legislators and attempting to persuade/bully tax administrators into doing long-term damage to the government’s negotiating position. Something to watch for.)

3. The Inflation Reduction Act

There is a pervasive narrative, supported by many talking points to this effect, that the Inflation Reduction Act (IRA) led to Direct File. And it did, but maybe not in the way you think. The IRA’s impact wasn’t just money, or a report; it completely inverted the political calculus around Direct File.

Certainly the IRA’s $80 billion in IRS funding provided ample elbow room for new initiatives like Direct File, and I never again had to answer the question, “But where will the money come from?” However, over the entirety of my time on the project, the IRS spent on Direct File an amount less than 0.1% of the funding provided by the IRA. As a share of the IRS’s scant enacted budget excluding IRA funds, Direct File would have comprised 0.26%, the equivalent of an ad-free Netflix subscription for a household earning $80,000. The IRA funds put wind in our sails, but Direct File was never going to break the bank.

The IRA contained a provision asking the IRS to spend $15 million to write a report about Direct File. After delivery of the report, the Secretary of the Treasury directed the IRS to pilot Direct File in filing season 2024, and the rest is history. But as the author of multiple sections of that report (full credit to Amy Paris as the primary author; she would agree), there’s no way our words alone would have been worth $15 million, particularly compared to the $24.6 million price tag of producing a report and shipping the damn thing (and producing a second report as a bonus). There’s a case to be made that the report to Congress was politically useful, but I think this case is overstated and relies heavily on post hoc reasoning. I can only hope lawmakers realize before repeating the experiment that despite Direct File’s success, the likeliest outcome of the provision was a costly but swiftly forgotten report. (Oh. Oh no.)

No, the IRA’s greatest contribution to Direct File: it passed.

During the early discussions of Direct File, the greatest reluctance came from a contingent, particularly in Treasury, who believed it would jeopardize the legislative chances of what was then called the Build Back Better Act, and in particular, its banner rejuvenation of the emaciated IRS budget. The Senate was split 50/50, and Democrats only held a technical majority via the tie-breaking vote of the Vice President. Unable to lose even a single vote, the camp resisting Direct File viewed movement on a “controversial” program as the straw that would break the camel from West Virginia’s back.^[8]

Even as Build Back Better’s prospects looked increasingly grim, hope sprang eternal on Pennsylvania Avenue,^[9] right up until Senator Manchin hammered what appeared to be the final nail in the act’s coffin on July 14, 2022. But less than two weeks later, well before this new reality could sink in, Senators Manchin and Schumer announced they had struck a secret deal to save the tax and climate portions of the bill, and unveiled the IRA. It swiftly passed and was signed into law on August 16.

Not only was their Direct File hesitance now obviated by the bill passing Congress, the proponents of additional IRS funding faced a new challenge. They needed to demonstrate to the American public the value of investing $80 billion in the IRS. For policy wonks, the case was a slam dunk. The IRA would enable the IRS to pursue high-income individuals and multinational corporations, whose complex returns had escaped scrutiny as the IRS was starved of resources. The Congressional Budget Office estimates that each additional dollar invested in tax enforcement yields between $5 and $9 of additional revenue. More funding for the IRS wasn’t spending money, it was raising money by enforcing the laws already on the books.

But $80 billion is a lot of money, and even the wonks understood that the IRS needed to show taxpayers a tangible return for this massive investment. Answering phones and clearing paper backlogs were obvious and important service wins, but these accomplishments wouldn’t be felt by most taxpayers.

The IRS needed something bigger, something exciting, something that showed that the bar had been raised for the kind of services taxpayers could expect from an adequately funded agency.

Consensus was within reach. Direct File was no longer a thing that couldn’t be changed, and it had quite suddenly become a change that voices across government were clamoring for, if only the IRS could find the courage.

• ← Previous
  Sittin’ On the Docket At VA