Michigan's Gotion fiasco proves subsidies can't replace a plan

What is a package of nearly $175 million in incentives worth if the rules change halfway through the game? Michigan found out the hard way. In October 2025 the state moved to claw back millions from a battery project that never really got started, and in our view it is one of the cleanest case studies of how not to run industrial policy.

A $2.36 billion promise ended as an empty field

According to CBS News Detroit, Gov. Gretchen Whitmer backed the $2.36 billion factory in 2022 and lawmakers approved nearly $175 million in incentives. The state then held Gotion Inc. in default of $23.6 million, saying no "eligible activities" had occurred on the site in over 120 days. Another $26.4 million of unspent grant money goes back to the state, and a separate $125 million grant was never distributed.

The plant near Big Rapids was meant to make cathodes and anodes and create over 2,000 jobs. Instead it drew years of resistance: Green Township voters recalled five local officials in 2023, and Rep. John Moolenaar pushed a bill, signed by President Trump, that stopped companies like Gotion from receiving EV tax credits under the Inflation Reduction Act.

Incentives buy a groundbreaking, not a business

The lazy reading is "subsidies failed." We think the more accurate one is that subsidies were asked to do a job they can't do. A grant can pay for land. It cannot create demand for the cathodes, calm a township, or guarantee that federal tax credits will still apply to your tenant a couple of years later.

Look at what the state is actually trying to recover. The $23.6 million that was spent went toward buying the site's land. That is the kind of outlay that looks like progress in a press release and leaves an empty field if the project dies. Meanwhile the $125 million grant was never handed over, with the state citing a lack of progress. That probably spared taxpayers far more than the clawback will recover. If there is a silver lining, it's the contract design, not the strategy.

And a clawback is a consolation prize. Even if every dollar comes back, the state spent years and political capital on a project that produced no cathodes, no anodes and, as far as the documentation shows, no jobs. Recovered money doesn't recover lost time, and time is exactly what states are supposedly buying when they compete for factories.

Policy whiplash is the real risk

The federal swing is the part Michigan could not control, and it matters. A project conceived under one administration's industrial logic ran into a very different one. CBS notes that auto manufacturing is the lifeblood of Michigan's economy, which has been hard hit by the tariffs imposed on the industry in 2025. That is a poor backdrop for a bet that depended on federal goodwill. The Institute for Energy Research, an advocacy group with its own perspective, reported in April 2025 that manufacturers had recently cancelled plans for $8 billion in proposed projects. Gotion isn't an isolated case, though this one has a distinctive political twist.

That twist is worth separating out. Some residents objected on environmental grounds, and Moolenaar, the project's loudest critic, accused Gotion of ties to forced labor and the Chinese Communist Party. Gotion did not respond to CBS's requests for comment, so those accusations remain just that in what we've seen. Whatever you think of each argument, a project that loses its community and its federal tax-credit eligibility has no stable foundation. No incentive schedule fixes that.

Governors should be pickier, not stingier

We think the takeaway isn't to stop competing for factories. It's to be pickier. Tie money to milestones, as Michigan at least partly did. Vet ownership and federal exposure before the announcement, not after. Win local consent early, before a recall election does the polling for you. And stop treating headline investment figures and job counts as results; the 2,000 jobs were never real until someone was hired.

The verdict: Michigan's clawback is a decent piece of damage control on a poorly conceived bet. Subsidy-led industrial policy works when demand and rules are predictable. When they aren't, a grant is just a very expensive option on a future nobody can promise.