They Were Right About the Problem. They Just Didn't Know What to Do About It.

Let's Give Them the Full Argument

I've spent two articles critiquing the current tariff structure on battery inputs, and in the interest of being taken seriously I want to start this one differently. The people who designed and defended the graphite tariff are not stupid. Some of their arguments are genuinely good. Before I dismantle them, I'm going to make them as well as I can.

This is the steelman. These are the strongest versions of the pro-tariff arguments on graphite and battery supply chain policy — presented without editorial commentary, as a defender of that position would make them. Read them charitably. Some of them will land.

That's the strongest version of the case. Several of those arguments are correct. Arguments 1 and 6 are genuinely compelling. Argument 3 has real economic logic behind it. Argument 5 lands, and I'll deal with it directly.

Now let's look at what was actually available to the administration that made these arguments — the technology that existed, the investment pathways that were open, and the timeline that was achievable within a single presidential term. Because this is where the case against them stops being about principles and starts being about competence.

What Was Actually Available

The administration's position, generously stated, was: China dominates graphite, that's a strategic problem, we need domestic production, and tariffs create the market signal to build it. Every word of that is defensible. What it required to be more than an argument — what it required to be a policy — was an answer to the next question: given the tariff, given the revenue, given the stated goal of domestic graphite production, what specific technology would we invest in, at what scale, on what timeline?

That question was never answered. Not because the answer didn't exist. It existed in exquisite technical detail, sitting in DOE laboratories, university research programs, and venture-backed startups, some of which had already won federal R&D awards. Here is what was available.

The Economics Before You Even Touch the Tariff Revenue

This matters: biographite from agricultural waste already has a modeled minimum selling price of $3.30/kg against a Chinese synthetic graphite market price of approximately $4.20/kg. That's a 27% profit margin at current market prices, before any tariff protection is applied. The domestic alternative doesn't need a 160% tariff to be viable. It needs capital to scale from a laboratory process to industrial production volume, and a customer — an OEM or a cell manufacturer willing to sign a multi-year offtake agreement that justifies the investment.

The government's role in that equation is not complicated. It's the same role it played in shale gas, in agricultural biofuels, in the semiconductor industry in the 1980s. De-risk the early capital. Guarantee the first customers. Get the first facility built. After that, the economics carry themselves.

What does that cost? The Flash Joule Heating process runs at approximately $500 per ton in electricity costs. A mid-scale graphite processing facility producing 50,000 tonnes per year — about 3% of current U.S. battery demand — requires roughly $200–400 million in capital investment. That is one aircraft carrier. That is less than one week of the interest on the federal debt. That is, as it happens, a fraction of the annual tariff revenue the graphite tariff alone generates.

The tariff collects over a billion dollars a year in revenue from the graphite problem. A fraction of that — less than one year's collection — would be sufficient to capitalize the first commercial-scale domestic graphite production facility using technology that already exists, already works at pilot scale, and is already cheaper than the Chinese import price before the tariff is applied.

That money is not being spent on domestic graphite. It is being absorbed into general revenue. The tariff exists. The revenue exists. The technology exists. The investment does not.

The Timeline That Was Achievable

The pro-tariff argument that "industrial policy takes a decade" is historically true in general and specifically false in this case. It applies to building a mining and processing industry from scratch. It does not apply to scaling a technology that is already at TRL 5–6, already cost-competitive, and already sitting in a federally funded national laboratory waiting for a commercialization partner.

Here is what was achievable within a single presidential term, starting January 2025:

Before the end of a single presidential term. Not a complete solution — 50,000 tonnes is about 3% of projected U.S. demand. But it is a running facility. It is domestic production. It is proof of concept at industrial scale. It is the foundation for the second facility, and the third. It is — to use the administration's own language — winning.

Now Let's Answer the IRA Objection

Argument 5 in the steelman is the most intellectually honest one, and it deserves a direct response. The IRA's FEOC provisions are functionally a tariff delivered through the tax code, and it's fair to ask why one deserves praise and the other condemnation.

The difference isn't the mechanism. It's the architecture. The IRA paired the FEOC exclusion with positive investment incentives — the 30D consumer credit, the 45X advanced manufacturing credit, the loan guarantee programs through the DOE Loan Programs Office. It applied pressure to Chinese supply chains while simultaneously funding the domestic alternative. Stick and carrot, deployed simultaneously, at meaningful scale.

Section 301 and the reciprocal tariffs applied only the stick. No funded domestic substitution pathway. No production incentives attached to the tariff revenue. No offtake guarantee to de-risk private capital. One policy created conditions for domestic industry to grow. The other created conditions for consumers to pay more while the industry waited for a capital signal that never came.

The IRA criticism in this series isn't that it was perfect. It's that gutting the IRA investment mechanisms while expanding the tariff stick is the worst of both worlds. You remove the carrot and keep the stick, which produces neither domestic production nor affordable vehicles. It produces only the stick.

The Verdict

Let's be precise about what the charge is. It is not that the pro-tariff argument was wrong. The diagnosis was correct: Chinese graphite dominance is a genuine strategic vulnerability, the supply chain concentration is dangerous, and domestic production capacity is a legitimate national security objective. Those arguments stand.

The charge is that the administration identified the right problem and then reached for the simplest possible tool — a tariff — without doing the work that would make the tariff functional as industrial policy. The technology to build a domestic graphite supply chain existed. Some of it was already funded by DOE. Some of it was already producing battery-grade material at pilot scale in American national laboratories. Some of it was already cost-competitive with Chinese imports before any tariff was applied. The timeline to first commercial production, with committed capital and offtake guarantees, was within a single presidential term.

None of that investment happened at the scale the problem required. The tariff revenue — over a billion dollars a year — went to the general fund. Universal Matter is still a startup. Solidion's ARPA-E grant is still a research project. The biomass biographite process that can produce anode-grade graphite for $3.30/kg from American agricultural waste is still waiting for its first commercial customer and its first committed capital partner.

The consumer is paying $685 per vehicle in graphite tariff. That revenue is going to the U.S. general fund — not to the Chinese exporter, and not to the domestic industry the tariff was supposed to build. The technology that would build that industry exists, is cheaper than the Chinese import it would replace, and is waiting for capital that isn't coming.

You can design a 160% tariff on graphite and call it industrial policy. You can stand at a podium and explain that China restricted graphite exports in 2023 and that domestic production is a national security imperative and that American workers deserve these jobs. Every sentence of that speech can be true.

But at some point, someone has to ask: given all of that — the tariff, the revenue, the stated goal, the available technology, the achievable timeline — what did you actually do with it?

The answer is in the BOM. It's in the $685 per vehicle that is still flowing to Chinese graphite exporters, tariff included. It's in the ARPA-E grant that should have been a construction loan. It's in the Rice University lab that built a process capable of turning American trash into battery-grade graphite for $500 a ton in electricity — and is still waiting for a government customer serious enough to scale it.

Industrial policy is not a speech about what you want to build. Industrial policy is building it.

The technology existed. The money existed. The timeline was achievable. The will to execute — the basic engineering discipline of stating a goal, identifying a pathway, allocating the resource, and following through — that's what didn't exist. You can't fix that with a tariff.