How has US industrial policy impacted climatetech investment?

The episode starts from a disagreement. On a panel prep call, another investor told Kann that a year after the IRA the impact was mostly theoretical; Kann’s own view was that multiple markets had already been transformed. Houser answers with a measurement rather than a position. Rhodium built the Clean Investment Monitor with a research center at MIT specifically to count what he calls steel on the ground rather than press releases, with a consistent series back to 2018 and a deliberately high bar for counting something even as an announcement: a specific site, a real construction timeline rather than a round-numbered 2030, and, for large projects, front-end engineering and design already underway, though not a final investment decision. On that basis the year from Q3 2022 through Q2 2023 saw $213 billion of clean energy investment broadly defined, up 37% year on year and 165% on five years earlier. On a Bureau of Economic Analysis definition, fixed structures and equipment plus durable consumer goods such as vehicles, that is 4.1% of all US investment against 1.7% five years ago.

Manufacturing is where the number surprises both of them: $39 billion in the past year, up 135%, against roughly $2 billion a year five years earlier, which Houser describes as very close to nothing. Kann’s objection is the obvious one: the IRA was barely a year old at the end of the data window, so the honest expectation would be site selection and state incentive negotiations, not construction. Houser’s answer is the hinge of the episode and the part most easily lost in summary. Much of this was not a response to a law signed in August 2022; it was companies moving on the expectation of one. The Biden administration had campaigned on climate, the rough outlines of the incentives were visible a year or more ahead, and firms took the bet. That is why Houser dates the policy-affected window to mid-2021 rather than to the IRA itself, and it makes the counterfactual messier than the headline implies. He is careful in both directions: with electric vehicle demand growing, some supply chain localization was coming regardless, because automakers prefer to build where they sell and parts of the chain want to sit near the customer. What the legislation changed was magnitude, not direction. Had it failed, he thinks many of those announcements would have fallen away. The investment is also narrow. Roughly 93% of the past two years of actual manufacturing investment sits in the electric vehicle value chain, meaning critical minerals, batteries, vehicle assembly and charging equipment, with most of the small remainder in solar.

The rest of the map sorts into three bins. Solar, storage and EV sales were already at escape velocity and are growing 15% to 40% a year; the legislation accelerates them, which matters for emissions because speed is the point, but it does not change their trajectory. The emerging technologies are the genuine creation story: carbon management, clean hydrogen and sustainable aviation fuel together drew $80 billion of announced investment over two years, a full third of announced wholesale deployment investment, for categories that barely existed before. But only $4.6 billion of that has become steel on the ground, the widest announcement-to-reality gap in the data, and for green hydrogen the reason is specific: projects are waiting on Treasury guidance before taking final investment decisions. The surprise inside that bin is blue hydrogen, where greenfield plants are being built now. Houser’s read is that they are penciling on 45Q, the carbon capture credit, rather than on the hydrogen production credit, because a project cannot claim both and 45Q’s policy pathway is far more predictable.

Then the third bin, where the incentives are losing. Wind investment and announcements have declined for two years despite an extension of the production tax credit out to 2032, because wind’s problems sit outside the generation economics. High-quality wind resource is less evenly distributed than solar, so wind is more exposed to transmission, siting and permitting constraints; offshore is slowed by permitting timelines and state procurement rules; and very large capital-intensive projects are unusually sensitive to high interest rates. Houser adds a comparative point that is easy to miss: wind could always claim the production tax credit, and the IRA let solar claim it too, which was a much larger change for solar than anything in the bill was for wind. His verdict is that the policy is necessary but not sufficient, and that without permitting reform and lower rates it will not produce the deployment growth required. Heat pumps tell a milder version of the same story. Installations were flat in investment terms and down slightly in units while continuing to take share from furnaces, because the whole residential HVAC market shrank: interest rates and households burning through pandemic savings hit residential construction and renovation hardest, and a $2,000 credit arriving into a market already squeezed by construction labor costs cannot outrun that. The pattern across both cases is the useful takeaway. Where a market’s constraint is something other than the economics of the technology, a tax credit does not reach it.