Loading issue research
Modernize federal tax administration around data-minimizing filing, tax-code neutrality between labor and automation, and a multilateral resolution to the digital services tax dispute.
Verification Status
AI-researched, unverifiedLast Reviewed
Jul 5, 2026
Cited Sources
15
Check how the claim was researched, how confident it is, and the evidence behind it.
This issue takes a position on three distinct places where tax policy and tax administration intersect with technology, joined by the fact that each is a place where the tax system already made a technology-relevant choice without examining it. On filing: revive a free, government-run, no-tracker filing channel for simple returns, and extend the same no-third-party-tracker floor to every IRS Authorized e-file Provider, reaching the paid commercial software most filers use rather than only the narrower Free File-certified set, so the privacy guarantee does not depend on adoption of the government option. On automation and the tax code: reject a flat tax on robots or automation as the wrong instrument, require Treasury and the Joint Committee on Taxation to disclose the effective tax-rate gap between labor and automation-capable equipment alongside any future depreciation or expensing legislation, and phase in an actual reduction of that gap over ten years by narrowing 2025's bonus depreciation and Section 179 expansion in step with an employer-side payroll tax cut sized to the disclosed gap. On digital services taxes: re-enter the multilateral OECD negotiation the United States exited in January 2025, rather than continue resolving each country's digital services tax through a bilateral tariff threat.
A tax system's privacy properties and its bias between labor and capital are structural, checkable features of its design, independent of any judgment about whether automation itself is desirable. A filing channel either does or does not let an advertising-funded company see a citizen's return before the government does; a depreciation schedule either does or does not tax an automated process more lightly than the paycheck it replaces. Both are measurable design choices, not verdicts on technology, and this issue's proposals target those design choices specifically, leaving broader judgments about automation, digital taxation, or private enterprise as categories to other issues in this platform.
Primary — Inclusive Growth and Economic Development. All three areas change who captures value the tax system currently gives away: ending the transfer of taxpayer data to advertising companies, correcting a tax-code subsidy that tilts investment toward automation at the margin regardless of which choice is more productive, and protecting U.S. digital firms' market access without conscripting unrelated exporters into the cost.
Secondary — Privacy, Security, and Trust. The filing proposals are PRIV-01's data-minimization principle applied to one specific, high-stakes point of contact between a citizen and the government: the one interaction where nearly every adult in the country hands over a detailed financial profile at once, currently to companies with a documented record of not protecting it.
On filing, the divide is partisan and specific. More than 150 House and Senate Democrats, led by Sens. Warren, Coons, and Wyden and Rep. Sherman, introduced the Direct File Act of 2026 to restore and permanently codify the program the Trump administration shut down; the same 2025 tax law that ended Direct File funded a $15 million Treasury task force to pursue free filing through expanded private-sector partnerships instead. This issue sides with a government-run channel, but for a narrower reason than either party states: Democrats argue government should provide the service because it's free and simple; this issue argues a government-run channel is the only one that can structurally guarantee no third-party data sharing, a guarantee no subsidized private alternative can make credibly given the industry's record. Proposal 2, the no-tracker mandate on certified private software, is where this issue agrees with the administration's private-sector-can-do-it framing, on the condition that the private sector be held to the standard it broke.
On automation and the tax code, neither party has taken up Acemoglu, Manera, and Restrepo's finding that the tax code itself, more than AI, is driving excess automation. Sen. Sanders' robot tax is the most visible current proposal and comes from the progressive wing of the Democratic caucus rather than the party's mainstream platform; the 2025 tax law's expansion of full expensing, passed with near-unanimous Republican support, moved policy in the opposite direction, widening the exact gap the tax-code research identifies. This issue rejects Sanders' instrument while accepting his diagnosis that current tax treatment of automation needs fixing, a synthesis neither party has assembled on its own.
On digital services taxes, both the Trump and Biden administrations agreed DSTs discriminate against U.S. firms; USTR made that finding in 2019 under Trump, and neither Biden nor Trump's second term reversed it. The parties differ on method: the Biden administration stayed at the OECD table through 2024; the current administration withdrew in January 2025 and moved to bilateral tariff threats. This issue's delta is a return to the abandoned method, rather than a new position on the underlying discrimination finding.
The strongest objection targets Proposal 1 specifically: reviving a government filing channel barely anyone used is a strange centerpiece for a privacy argument. TIGTA's own numbers are unflattering. About 751,000 of the roughly 30 million eligible taxpayers registered for Direct File in its final season, and only about 41% of those filed through it. A critic could reasonably argue that a data-minimization case built around a tool that reached under 3% of its eligible population is solving a problem almost no one encountered, while the pixel-tracking data-sharing problem was happening in the same tax season to tens of millions of people using the commercial software Direct File barely dented.
That's a fair reading of Direct File's own reach, and it's answered by what the rest of this issue does, rather than by defending Direct File's adoption curve. Proposal 2 doesn't route the privacy fix through getting people to switch products. It applies the no-third-party-tracker floor directly to the commercial software the other 97% of filers already use, the same software the 2023 congressional investigation caught sharing data. Reviving Direct File and mandating data minimization on private software are aimed at two different populations by design: the second proposal carries the privacy guarantee at scale, while Direct File guarantees an option with no commercial relationship at all, for filers who want one.
What this issue doesn't yet resolve is enforcement architecture: mandating a no-tracker standard on an industry that has already been caught routing around disclosure rules, including hiding its own free-filing page from search engines, requires an agency with audit authority and a penalty schedule with teeth, and the likeliest vehicle, the FTC's existing authority over deceptive practices, hasn't been tested against a data-minimization mandate of this specific kind. That's a gap in implementation detail, not in the case for requiring the mandate in the first place.
The commercial tax-prep industry bears the most concentrated and least sympathetic cost. A no-tracker mandate removes a monetization channel, behavioral ad data, that a 2023 congressional investigation already found being used in a way its own authors called likely illegal, and a revived free public option competes directly with simple-return products that currently generate revenue for Intuit, H&R Block, and their competitors. This issue accepts that cost without much hesitation: a company doesn't have a legitimate claim to revenue built on a data practice a group of sitting senators has already flagged as unlawful.
Federal taxpayers bear a modest, ongoing appropriation cost to run a Direct File-style channel, on the order of the $16 million to $61 million range TIGTA and the IRS have each separately estimated for a single recent fiscal year, spread across the entire tax base rather than concentrated on any one group. That cost is comparable to, and by TIGTA's corrected accounting likely smaller than, what the government was already spending to run the pilot it just shut down, which makes it a modest price for closing a documented privacy gap.
The largest U.S. digital multinationals, Google, Meta, Apple, Amazon, and a handful of others, bear a strategic cost from this issue's digital-tax proposal specifically: re-engaging a Pillar One-style negotiation means accepting a binding, permanent formula for how much tax authority market countries get over their profits, in exchange for ending the current patchwork. That could cost these firms more in aggregate than continuing to use tariff threats to suppress DSTs one country at a time while the underlying multilateral question stays unresolved indefinitely. This issue accepts that trade because the alternative has already shown its own cost: French cheesemakers and champagne producers were threatened with a 25% tariff in a fight about companies whose names never appeared on the tariff list, a threat suspended before taking effect but still a real, uncompensated compliance and uncertainty cost for exporters with no stake in the underlying digital-tax dispute, and that collateral pattern repeats every time the fight is refought bilaterally instead of settled once.
Capital-intensive firms that just secured permanent 100% bonus depreciation and an expanded Section 179 cap in 2025 bear the direct cost of the phase-in: a preference they were told was permanent narrows every year for a decade, on a schedule tied to a number, JCT's own gap estimate, that neither party has previously used to size a tax change. That cost is real and concentrated, not diffuse, on manufacturers, logistics operators, and other equipment-heavy sectors that lobbied hard for the 2025 provisions. This issue accepts it because the alternative is a tax code that keeps telling a firm at the margin that a machine costs less in tax terms than a hire for the identical task, a subsidy no one voted for as automation policy, only as investment policy.
Turn frustration into useful pressure.
If this position misses evidence or a lived consequence, challenge it. If it holds up, help test it locally and connect it to the issues around it.