The $41 million threat to the middleman: Why the "efficiency" committee really killed IRS Direct File

The Internal Revenue Service spent $41 million to build a functional piece of consumer software. It shipped on time, operated securely, and closed its pilot phase with a staggering 94% user satisfaction rate. In the historically cursed realm of federal IT procurement, this is practically a miracle. And that is exactly why the newly minted Department of Government Efficiency and its tech-elite backers decided it had to die.

If you ask the libertarian techno-optimists currently steering federal restructuring efforts, the government is fundamentally incapable of building modern digital infrastructure. I am willing to grant them some heavy ground here. Federal software projects do indeed have an agonizing history of becoming bloated, over-budget clusterfucks. We all remember the catastrophic rollout of Healthcare.gov, and anyone who has attempted to navigate a state unemployment portal or renew a business license online knows those platforms are usually held together by obsolete mainframes and prayers. When Silicon Valley executives argue that federal agencies struggle to ship consumer-grade technology, they are pointing at a very real, well-documented problem.

But the IRS Direct File pilot completely broke that mold. It was a lean, targeted deployment that operated flawlessly.

I spent a couple of hours this week reading the UC Berkeley Direct File report to understand exactly what the efficiency committee found so legally and morally objectionable. When you filter out the political noise and just read the unglamorous, 23-page post-mortem, the reality of the program becomes entirely undeniable. The software was not built on blockchain vaporware or generative AI hype. It did not require millions of dollars in consulting fees paid to private firms to produce vague slide decks about synergy. It was built by, as the report explicitly puts it, "determined leaders delivering world-class digital services."

Let us look at the actual ledger of this project. The official IRS pilot data confirms an incredibly successful 2024 run. For a total budget of $41 million—roughly the price of a mid-tier, venture-backed crypto scam before the founders inevitably flee to a non-extradition treaty country—the agency deployed a system that allowed hundreds of thousands of taxpayers across twelve states to file their returns directly with the federal government. It was entirely free. It was available in English and Spanish. Users saved an estimated $5.6 million in tax preparation fees and completely avoided the dark-pattern interfaces designed to upsell them on unnecessary audit defense packages.

The narrative currently being peddled by the efficiency committee is that shuttering this program is a vital, urgent step to curb runaway bureaucratic waste. They want the public to believe that Direct File was just another bloated agency boondoggle draining the treasury, framing the platform as an unnecessary disgorgement of taxpayer funds that must be aggressively amputated to save the republic.

That is entirely fabricated.

If shipping a universally beloved, working federal tax portal for a fraction of what the Pentagon spends misplacing spare aviation parts is categorized as waste, what exactly is efficiency? The committee is not demanding a leaner government; they are demanding a stupider one.

To understand why a highly functional, universally liked public utility is being actively dismantled, you simply have to follow the forensic ledger of corporate lobbying. For over two decades, Intuit has waged a sophisticated, sometimes covert war to maintain the Free File Alliance non-compete agreement. The basic business model of the American tax preparation industry relies entirely on preserving an artificial labyrinth of regulatory complexity. The government already possesses the W-2 data and knows exactly what you owe, but you are legally mandated to pay a private company to calculate that exact same number and hand it back to them.

When the Direct File pilot successfully proved that the federal government could simply bypass the middlemen and interact directly with citizens, Intuit executives panicked. A quick glance at the millions spent by Intuit on lobbying tells the entire, damning story. You can track the massive spikes in their federal expenditures correlating exactly with moments when public filing alternatives gain legislative traction. Their lobbying disclosures reveal an unrelenting pressure campaign deployed across multiple administrations, specifically targeting any legislative effort that might simplify the tax code or provide a free, government-backed filing option. They did not mobilize an army of K Street operatives because the IRS was being grossly inefficient. They mobilized because the IRS was finally being dangerously competent. The Direct File program was an existential threat to a highly lucrative, rent-seeking tollbooth, so the tollbooth operators bought enough political leverage to have the new road demolished.

The actual tech sector—the software engineers, security researchers, and infrastructure developers who build these systems for a living—understands exactly what is happening here. Scrolling through a recent discussion among tech workers analyzing the shutdown, the consensus from the people who actually know how to ship code is overwhelmingly clear. These are professionals who look at the underlying architecture, evaluate the deployment strategy, and recognize a spectacularly successful software launch when they see one.

They see the utter absurdity of a system that mandates you engage a private, for-profit corporation to submit mandatory financial data to a public agency (a deeply unserious shitshow, frankly). Direct File proved that we do not actually need Intuit's predatory architecture to fulfill our basic civic obligations. The true inefficiency targeted by the newly appointed bureaucratic overlords was the public realization that a better, entirely free system is technically trivial to maintain.

It is profoundly exhausting to sit at a desk and watch billionaires use the immense, opaque machinery of the state to crush a functional public good simply to protect a corporate donor's profit margin. The efficiency committee is actively destroying the most competent piece of government software shipped in a decade, and they are demanding we thank them for the privilege of being forced back into the arms of a private monopoly.

But we cannot afford to shrug and accept the dystopian premise that the internet is destined to be a series of private, rent-seeking tollbooths. Surrendering to the idea that the government is inherently incapable of serving the public only guarantees that it never will. The $41 million IRS Direct File pilot was a definitive proof of concept. Public digital infrastructure is possible, we just saw the undeniable receipts, and now we have to fight like hell to keep it.