Is the Inflation Reduction Act, passed nearly two years ago, doing what it set out to do? In this episode, Trevor Houser of the Rhodium Group compares the predictions of pre-IRA energy-sector models to the real-world data on clean-energy investment since its passage.
Text transcript:
David Roberts
During the previous US brush with climate policy, back around 2010 when the Waxman-Markey bill was being discussed, energy-system modeling tools were too slow and too crude to contribute much. Without the ability to do rapid, on-the-fly modeling of various policy options, it was difficult to say with any confidence what effects those policies would have. Consequently, the public debate was based almost entirely on vibes. (It did not go well.)
It's different these days. Energy-sector modeling has progressed in leaps and bounds and is now capable of almost real-time analysis of policy proposals. That meant the dialogue among policymakers and the public in the run-up to passage of the Inflation Reduction Act was uncharacteristically informed. We knew, to a decent approximation, what effect different options would have on greenhouse gas emissions.
We also have, here on the backside of IRA, an unusual profusion of information about what has happened in the economy in its wake, thanks in part to the creation of the Clean Investment Monitor, a collaboration between analysts at the Rhodium Group and MIT. It closely tracks public and private investment in every sector of the clean energy economy.
So, we had models beforehand to predict what would happen. We have almost two years of data about what did happen. Do they match? Were the modeling predictions correct? Is IRA doing what its backers said it would?
To find out I contacted Trevor Houser, a partner and lead analyst at Rhodium. We talked about which sectors have seen the predicted successes and which haven't, what kind of trends emerge from the investment data, and how IRA's performance to date is informing Rhodium’s subsequent projections.
All right then, with no further ado, Trevor Houser of the Rhodium Group. It is so great to finally get you on the podcast, man.
Trevor Houser
Thanks, Dave. It's great to be here.
David Roberts
It's been a long time coming. I'm excited for this. It's got a lot to talk about today. But I just will say I've been following your work since you were just a little baby analyst and I was just a little baby journalist.
Trevor Houser
I still remember the day I was doing my very first congressional testimony, terrified, to the House Select Committee on the Climate Crisis, on the Recovery Act and some very early energy system modeling, trying to analyze the effect of different provisions of ARRA. And this wild-hair, bearded reporter came running up to me after the session, and a friendship was born.
David Roberts
Beautiful. It's a beautiful story. And now here we are, saying something passed. Wacky.
Trevor Houser
Wacky.
David Roberts
Okay, so let's start by just talking a little bit about the Clean Investment Monitor. What is that? How did it come about and what is it for? Why did it get set up?
Trevor Houser
Yeah, so the goal with the Clean Investment Monitor is — over the past few years, there's been, as your listeners know well, some of the most consequential energy and climate legislation passed in US history: the combination of the Inflation Reduction Act and the Infrastructure Investment and Jobs Act, or otherwise known as the Bipartisan Infrastructure Law, that include long-term incentives and grants for clean energy technology, clean transportation, and other emission-reducing investments. And there were modeled projections in the run-up to the passage of the IIJA and the IRA by Rhodium and other groups, and then analysis right after the passage about what the potential impact of that legislation would be. And the Clean Investment Monitor, the goal is to provide as close to real-time as possible data on how that investment is going, what sectors, what technologies are attracting the most investment, which federal funding programs are moving quickly, which are not, what parts of the country that investment is occurring in, whether it's having its intended effects in terms of community economic development, to provide that type of information to stakeholders and policymakers so that they can identify where things are going well, where it's not, where more work is needed.
David Roberts
Right. It's a little weird to me that something like this didn't already exist. Was this all just public information that you're just kind of gathering and putting in a clear format?
Trevor Houser
Yeah. So it's collecting a lot of different existing data sources and harmonizing them and putting them into one place. But then there's also this very labor-intensive process of — it's a team of analysts that are combing through company filings, news reports, to capture announcements on investments in new manufacturing facilities, let's say. But then also to track, like, have they broken ground on that facility? How much are they investing this year versus last year in that facility? How quickly is it ramping up production? So we benefit from some existing data sets that just need to be pulled together and transformed and harmonized.
But then there's a lot of extremely hardworking research analysts that are combing through Google News alerts and reading mind-numbing 10-K filings from publicly traded companies.
David Roberts
Oh, no. I just have a couple of general questions about modeling. Before we get into the specifics. I'm just a little bit curious. I mean, like I said in the intro, it was a little bit of a new thing in the world to have policy being discussed and crafted in the presence of what is basically real-time modeling. I mean, so sort of abundance of information about what would happen. I wonder if you had any sort of impressions of how that changed the policy-making process or how it differed from other policy-making processes you've seen, I want to say sort of in retrospect, that the whole thing worked out pretty well.
That's kind of how you want things to go. I mean, obviously, we'd like the politics to be better, but in terms of informed debate, it seemed like weirdly informed.
Trevor Houser
Yeah, it's a great question. Maybe let me answer it just from a Rhodium lens. And what our journey has been on this because as we talked about a minute ago, we did our first energy system modeling to try to inform a live fire policy exercise with Waxman-Markey and then later with the Recovery Act. And over time have been refining and evolving both our modeling tools and the way that we engage policymakers and stakeholders to figure out how to make it more useful and to learn from past mistakes. And in other modeling groups I've been at, I think I've been on the same journey.
And a couple of the things that we have learned is that, so, one, there's things you need to do to the tools just to be able to run them more quickly, to be responsive on timescale.
David Roberts
Is that something about the tools themselves, or is it just more computing power that allows this new speed?
Trevor Houser
So, there's a portion of this about the tools. It's about building the models and computational platforms that let you run them quickly. But it's also just about prep. So, if you think about the IRA, about two years before the passage of the IRA, our team at Rhodium started thinking through possible electoral outcomes in the 2020 election and gaming out in those electoral outcomes what types of policy proposals might be put forward and began speculatively modeling them to work through some of the places where the models weren't working well, where we would need to make improvements to them.
You have to do a lot of that work in advance if you want to be able to respond quickly to policy developments as they emerge. You kind of need to have a library of potential policies that you've worked through the kinks of modeling so that when you get the specifics about, "okay, this tax credit is going to be extended for this period of time at this level," you can do that analysis much quicker than if the very first time you're trying to figure out how to model a tax credit in your model is when it's being proposed in Congress you're going to be a year too late.
David Roberts
And you found lawmakers sort of eager for this, like, eager for the new information. You didn't have to fight to push this into the process, in other words?
Trevor Houser
No, I think both legislative staff and administration officials and key stakeholders — environmental groups, unions, industry associations, others — were pretty desperate for objective, independent analysis. And in part, because climate is a fundamentally quantitative exercise, it comes down to this unit of measure of a ton, and against a backdrop of a very specific US target, under the Paris Agreement of a 50% to 52% reduction, a top of mind question for all of those stakeholders was, "Okay, how far does it get us? How do we design a policy that is designed in a way that it can pass, but also does as much as possible to move us towards that target?" And I think particularly with the Inflation Reduction Act, the incredible congressional staff that were working nights and weekends for years to put together that legislation, also were looking for independent analysis to just anchor their own thinking relative to what they hear from stakeholder groups around any particular piece.
Right, because everybody has a particular component that they care about a lot. And if you're drafting that kind of legislation, you want to know, what does it all add up to? What matters more from an emissions standpoint?
David Roberts
It's just worth noting. I think everybody going through this process of policy formation, working from the same information about what it would do is just, like, a little bit crazy. It's a little bit of a new thing in the world. And just a final modeling question. And this is a very basic question, but I think just like, if you're kind of a normie out in the world and someone brings this model to you and says, "hey, we have a model, and here's what it shows," you sort of, like, just have to accept it, but there's no real magic here.
So just say a little bit about — so, you're like, modeling what these tax credits will do to EV sales, for instance. Just sort of like, what are the types of information that go into making that calculation?
Trevor Houser
Yeah. Like, what's the mechanics in the model for it?
David Roberts
Yeah, what are the mechanics of doing that?
Trevor Houser
Yeah. So you need to have a bunch of information about what vehicles cost and what they'll cost in the future. Now, in areas where there's a lot of uncertainty about what vehicles will cost in the future, like, "what will the price of lithium-ion batteries be?" a kind of prudent approach to modeling, one that we take is you use a range of potential future changes so that you are reasonably confident that your projections are capturing the full range of likely real-world outcomes. Then you need to assess consumer behavior: how responsive will customers be to a tax credit?
What range of EVs are available? Are there going to be enough models available that they'll satisfy the full range of consumer preferences? Right. What's the price of gasoline going to be? So, when someone is deciding between an EV or an internal combustion engine, they're comparing the price of electricity to the price of gasoline. So, you need to know what both of those things will be. And again, it's impossible to predict specifically what any one of those inputs will look like 5-10 years from now. So, you try to develop reasonable ranges using research from other groups as well as your own team's judgment. And then you build a model, or operate a model that can capture the interaction of those factors.
So in the total cost of ownership of an electric vehicle, how much of it is the battery price? How much of it is the cost of electricity compared to an internal combustion engine vehicle? How much is the cost of gasoline versus the cost of repairs versus the upfront cost of the car?
David Roberts
And so there are some judgment calls and some predictions made. Because I just want to frame this: so, like, if the actual results are off from the models, it's not that the policy was bad or wrong or that the model was bad or wrong. It's just like you're predicting a lot of future developments, and prediction is hard.
Trevor Houser
And it's important as a modeler to regularly go back and look — and do so bravely — at what did we get right, what did we get wrong? Right. And that's one thing that we're trying to do with the Clean Investment Monitor. As we get real-time information of what's actually happening, compare that to the model projections and ask ourselves, like, "Is that close to what was modeled?" If not, why? Was it a wrong assumption in the modeling, or has the world changed in a way that we couldn't predict?
David Roberts
Right. I want to get into some of that later, but, like, one example that I kind of wanted to ask about specifically here is interest rates. Because every discussion I have practically these days, one way or the other, comes back to interest rates and their unwelcome effects on investment. Did you predict the course of interest rates, and how big of an effect does that have on the results?
Trevor Houser
I mean, if you go back to like, 2020 or early 2021, our projections did not have a large increase in interest rates. Now, we update our core modeling every year through our Taking Stock report for the US, which is the baseline that we then use for policy analysis. As we got into 2021, 2022, 2023, and interest rates were rising, then that starts showing up in your projections as well. But, yeah, that type of macro shock is exactly the type of thing that's pretty challenging to capture in a model. One important thing to flag when modeling a policy is usually what you're trying to focus policymakers on in your modeling is not so much the absolute outcome like this thing will be at this point in this year as much as what is the delta relative to a baseline from the policy.
And you're including the same assumptions both in your baseline and in the delta. Right. So, for example, in assessing the impact of the IRA, which by 2030 was between a 700 million metric ton and a one gigaton a year intervention, that delta, the interest rate assumptions are both in the baseline and in the policy scenario. Right. And so, if there's a shock, an unforeseen shock, that makes the point prediction in the future wrong, it's less likely to change the delta between the baseline and the policy future. Does that make sense?
David Roberts
They both shift in the same direction.
Trevor Houser
They both shift in the same direction. So a high interest rate future means that the non-IRA baseline would have been much different than the baseline against which we and other modeling groups were analyzing the impact of the IRA.
David Roberts
That's if you can get policymakers to interpret models the right way, which is no small thing.
Trevor Houser
No small feat.
David Roberts
So let's talk then about, so, we have, what, a year and a half of data, post-IRA data? That's what we're working with here. So let's start first with kind of the big overall number, sort of like how much private investment in clean energy, broadly, did IRA pull forward, and how do we know how much of that IRA is responsible for?
Trevor Houser
Okay, so if you look at 2023, there was $239 billion in investment in clean energy and transportation. So, in the Clean Investment Monitor, we define that as investment in manufacturing, clean energy technologies and clean vehicles, investment in the deployment of those clean energy technologies and vehicles, both in the wholesale energy market and in the retail sector. So, $239 billion. That was up 38% year on year. And if you compare it to 2021, there was $139 billion in clean energy investment in the US. And in 2023, it was $239 billion. So 80% plus increase over that two-year period in clean energy investment in the US. And that's adjusted for inflation.
David Roberts
Now, presumably the arc, the line, was going up and to the right already.
Trevor Houser
Yes. So how much was IRA responsible?
David Roberts
And is there a hard and fast answer to that question?
Trevor Houser
Yeah. So we don't know for sure. It's very difficult to do that type of attribution analysis. I mean, we can tell you what the models predicted the delta would be, but it's hard to say in reality. Qualitatively, what we see when you look under the hood at the data. I would break it into three categories. So there's a category of technologies for which I think it's pretty safe to say that IRA and the bipartisan infrastructure bill are really responsible for the majority of the investment, that those incentives fundamentally change the economics. Right.
So I would put in that category most of the clean energy manufacturing, the EV supply chain, battery supply chain, solar. That's really been, I think, to date, the most successful part of the Inflation Reduction Act is just how much it's catalyzing clean energy manufacturing investment and in lots of different technology segments. And then the other would be an emerging climate technologies. So hydrogen and carbon management.
David Roberts
We'll touch on those later, too.
Trevor Houser
Yeah, so those two. Those are the ones that are really growing exponentially. And if you look at the incentives provided in the IRA, they really are transformative in terms of the overall economics. That's been one category. Two are technologies that were already on and had achieved escape velocity. They're already on that upswing, and the incentives in the IRA are accelerating that. But it's hard to know exactly how much. Right.
So I would put in that bin, utility-scale solar and storage, that account for the majority of the utility-scale clean energy investment that we're seeing. And then there's a third category, which are technologies that where investment's declining despite the IRA. Now, it's entirely possible that without the IRA, they would be declining faster, but they're really struggling. The incentives in the IRA have not been able to overcome the other headwinds. And that's primarily wind and heat pumps. And we can talk about those in more detail, too.
David Roberts
So we can say, I mean, intuitively, you want to say they'd probably be doing even worse without the incentives.
Trevor Houser
Yeah, directionally, that has to be true. The question is just the magnitude.
David Roberts
Right. And then also what your recent report did, which was interesting, is tease out how much government money is being spent, how much public money is being spent, because the whole policy premise of IRA is that government investment catalyzes private investment. Right. Pulls in private investment. So there's $239 billion private investment relative to how much public spending.
Trevor Houser
$239 billion is the total investment. The federal spending would be a portion of that. So, unfortunately, to do this comparison, we only have the federal investment for fiscal year 2023. So that runs through September 30th. So we've got to look at a slightly different time horizon, but it'll give you a sense of the ratio. So in fiscal year, government, fiscal year 2023, so that would have been October 1st, 2022, through September 30th, 2023. Right? In that period, there was $220 billion in total investment. Of that, $34 billion is federal investment. So it's a five to six ratio between the public and the private.
David Roberts
What do we make of that? What do you learn from that? What conclusions can you draw from that ratio?
Trevor Houser
Yeah, so first, it is clear that the vast majority of the investment required to decarbonize the US is going to be private. It's not going to come from the federal government. And the federal government has a very powerful role to play in providing incentives that shape the direction of that investment in a low carbon way. And so far, those investments, the majority of which are tax credits. So one really unique thing that we're doing at the Clean Investment Monitor is project by project, estimating what the tax credit uptake is likely to be and then rolling that up, because while we get government reporting on grants and loans, we don't have data on tax credit uptake. So we need to estimate it.
David Roberts
So, presumably, that's private information to the company. So, that's not made public anywhere.
Trevor Houser
Exactly. It's in their IRS filings, but because of confidentiality, the government can't — in a couple of years, they will probably release aggregate estimates. But you don't get the kind of timely and detailed information that you need from government sources. And so, we have to estimate it. From the bottom up. Right.
David Roberts
And so, I mean, it seems like, on its face, that if private spending is, what did you say, six times public spending, that that is working, right? Seems like good news. Seems like public capital is, in fact, catalyzing large amounts of private capital.
Trevor Houser
We think it's a pretty good ratio. I think it's a pretty good ratio. It's a pretty good, pretty good deal for taxpayers.
David Roberts
Was the ratio in question? You know what I mean? Did you have questions in your mind about what that ratio was going to be?
Trevor Houser
Yeah, and I think, in particular, in some ways with tax credits, some of the ratio you kind of know, because it's a tax credit that's covering 30% of investment. Right. Or 10% of the cost of production. So that ratio should be relatively constant. A question is just the magnitude, like, how much investment actually happens once those incentives are in play. But, I think that's for those of us who are wading through the details of the tax code provisions. I think for the general public, there is a perception, or can be a perception, that this type of investment activity is solely being paid for by the taxpayer. Right.
David Roberts
Yes. And that's certainly how Republicans talk about it, right? Like, what is the cost of the bill, et cetera, et cetera, et cetera.
Trevor Houser
Right. And so you can look at what is the fiscal cost of the bill, and you need to balance against that what are the benefits of the bill, both in terms of the emissions that it's reducing, but also the private sector investment that it's catalyzing and associated job creation and local economic development activity.
David Roberts
Right. And IRA was a lot of things, a lot of different things, only some of which have started to actually spend money. Right. So do you have a good global sense of sort of the money spending provisions in IRA? How much are actually spending money? Like, presumably we're going to see that public amount go up as all the spending programs come online. Is that fair?
Trevor Houser
Yeah, though less than you would think, because the majority of the federal investment under the IRA is tax credits. And the nice thing about tax credits is that they are effective pretty much immediately. You don't have to have a team at an agency that is taking proposals and allocating grants. You change the law and literally the next day companies can start claiming the credits on their tax returns. Now, for some of the new tax credits that were created — so, the IRA, for some tax credits, they existed before and the IRA extended them. And for those companies know how to take advantage of those credits, the rules weren't really changing.
That was pretty straightforward. The effect of those was pretty immediate. For other credits that were new in the IRA, the Treasury Department and the IRS had to develop guidance on how to interpret the legislative language in the IRA. And so for hydrogen, for example, for green hydrogen, the 45V tax credit, that guidance didn't come until late in the year. And so as a result, there hasn't been much tax credit adoption under 45V because industry was waiting on the guidance to be issued by the Treasury Department, whereas the EV purchase credit and the clean electricity tax credits, those were increases and extensions of credits that already existed and so they moved much quicker into the economy.
Now, going forward, we'll see of that $34 billion in federal investment, only about $400 million of that was grants and loans. Now, important caveat: loans, here we're looking at the fiscal cost of the loan, which is not the total amount of the loan, it's how much that loan costs taxpayers in terms of the subsidy provided. So of the $34 billion, only a small fraction to date has been grants and loans. That loan and grant number will grow going forward because those programs take a little bit longer to stand up.
David Roberts
So let's get to the meat of the thing. Let's talk about some sectors and what we found out. So let's start with EVs. Rhodium did a report basically honing in on two sectors, EVs and electricity. How did they perform relative to what the models said? And the story for EVs is positive. Basically, they're on track to what you expected. Is that more or less, right?
Trevor Houser
Yeah. Despite what you would have read in the news for the past three months.
David Roberts
I know. Despite the cavalcade of negative stories and doom saying they're on track, doing as expected, doing what policymakers thought they would do, doing what you projected, more or less.
Trevor Houser
Yep. And so this is both Rhodium and then collaborating with us on this was the REPEAT project at Princeton and in Energy Innovation. So those were two of the other groups that provided a lot of the modeling before and after the IRA. So we all got together to do the brave thing of "Let's compare the real world to what we said was going to happen. And how does it look?"
David Roberts
Held hands and jumped in together.
Trevor Houser
And jumped in together. Jumped in together. And, yeah, for EVs in 2023, EVs accounted — was ZEV, so it includes plug-in hybrids, battery electric vehicles, and then a tiny number of fuel cell vehicles, it's like 13 Toyota Mirais sold in California. So the zero-emission vehicles accounted for 9.2% of all passenger vehicle sales. And that is at the top of the range of the modeling groups had projected.
David Roberts
I had thought that the strict domestic content requirements, et cetera, that Manchin insisted on these, which restricted the pool of EVs that they applied to at the beginning, would slow things down. And then you got interest rates. I mean, it seems like all these negative stories were inspired by true events, as they say. So what explains EV's good performance here?
Trevor Houser
So, I think it's a couple of things. One is that this is a place where the delay in issuing guidance for the tax credits actually benefited sales, because the guidance, the implementation of those domestic requirements, it took a little while for the Treasury Department to finalize those regulations.
David Roberts
Oh, and were the old rules in place while they were working out the new rules?
Trevor Houser
It was like a hybrid, so it was an extension at the new level, but there was a transition period where there was a greater number of models that qualified. So that's one. Second is that a lot of those requirements only apply to purchased vehicles, not to fleet vehicles. Right. So they apply to personal purchased vehicles, but not fleet vehicles. And so if you lease an electric vehicle, often the credit is available independent of some of those restrictions. And there's a much larger share of leased vehicles over the past couple of years than there had been previously in the EV segment.
David Roberts
And PHEVs, plug-ins, kind of came on stronger than expected. Are they a bigger part of the ZEV portion than you expected?
Trevor Houser
So they grew a little bit quicker than expected. Their total numbers are fairly small. The market for traditional hybrids is much larger. The plug-in hybrids, they are growing faster than expected, but the total numbers are pretty small. So it's still the battery electric vehicle sales that drive the headline numbers.
David Roberts
And, is there a tidy story to tell about how the results in the EV space so far are informing your subsequent projections? I mean, I guess this just says, "yeah, your modeling is right on." So you're just going to keep the modeling in place. Are you tweaking anything? Is there anything you've learned that's tweaking your expectations about subsequent EV sales?"
Trevor Houser
Yeah, I mean, as part of the model updates we do each year, we update battery prices, vehicle costs, the number of models that are available, and we'll have the updated guidance. So, I don't know yet how that's going to net out in terms of whether our projections for EV sales are higher or lower for future years than they were previously, because there's a lot of different changing factors, many of which work at cross purposes. But yet, we look at what happened in 2023, how many vehicles were leased versus bought, what was the range of vehicles, and then looking at the model list in future years, what new models are coming online and when, and then updated projections from NREL or other groups about lithium-ion battery trends to try to shape the forecast for the next few years.
David Roberts
Right. And maybe LFP trends, maybe even sodium-ion trends. Who knows what could happen? So, EVs are on track. This was somewhat of a surprise, given the doomy vibes around EVs. In contrast, the vibes around clean electricity tend to be pretty good. Everybody loves solar and wind. They're booming, they're cheap, everybody keeps talking about it all the time. But electricity, clean electricity underperformed relative to model projections. Why? And by how much, I should say.
Trevor Houser
Yeah, so we installed a record 33 gigawatts of utility-scale renewables and storage in 2023. So, there was a ton of investment, a record amount of investment, but that came in at a little bit below the bottom end of the range of projections for the year.
David Roberts
So, this is not a catastrophic falling short. It's just a nagging more than you would like falling short.
Trevor Houser
It's like a flashing yellow light. It's like, "hey, there might be a problem here." Like, this needs some attention. Things are going well. Investment is growing a lot, but the pace of investment that we're going to need to deliver, the 40% emission reductions projected under the IRA, let alone the 50% to 52% that the US needs to hit its NDC, is just very high, very high. While the growth is fast, it is not fast enough. And so the question is why? And there are a number of different factors that we know to be at play. It's still a little bit difficult to tell the relative weight of those factors.
It kind of varies by technology, varies depending on which project developer you talk to, which equipment company you talk to. But let's run through those factors that are at play. So, the first are some, what we hope will be transitory factors, which are that change the economics in 2023, but that should alleviate over time. First and foremost is high interest rates. Right. Clean energy is much more capital intensive than fossil electricity, and so it's much more vulnerable to interest rate.
David Roberts
Do we, does anyone anticipate rates going down? I mean, not going up more? I think we can expect now, but do we expect them to go down?
Trevor Houser
The futures markets, for whatever they're worth, are currently betting that the Fed will start cutting rates in the summer.
David Roberts
Interesting.
Trevor Houser
And I think the expectations are that the rate cuts will be gradual, not fast, so that they will start coming down, but will be coming down pretty gradually over a multi-year period.
David Roberts
Right. So that was mainly, it seems like that mainly hit wind, is that right?
Trevor Houser
It disproportionately hits wind, but it hits solar and storage, too. It hits wind a little bit harder because the project timelines are longer. So, you're carrying your construction finance longer before you start generating revenue. But wind, solar storage, they're all capital, basically. There's no fuel cost. Right. So, interest rates hurt all of them.
David Roberts
So, that's one that's transitory.
Trevor Houser
Yeah. Supply chain issues that we believe to be transitory too where that — both the cost and availability of a lot of inputted goods was hard, whether that's solar tracking systems or transformers or particular wind components. That's an issue that was going to exist anyway when you're trying to quickly ramp up the amount of investment in installation you're doing, but was compounded by COVID-driven supply chain entanglement that's just taken a little bit longer to unwind. Those are the two transitory, or hopefully transitory issues. The question is how quickly they go away, but they should be transitory. And then there's a list of potentially more structural challenges that are going to need to get unwound.
And these are all what we've described as, like, non-cost barriers. So, the IRA successfully makes clean electricity, you know, temporary inflation notwithstanding, makes clean electricity and clean transportation cost-competitive with fossil alternatives pretty much everywhere in the country and in every application. But there are other barriers to the speed at which companies can realize that new cost parity in building new infrastructure and equipment. And this ranges from sighting and permitting issues and timelines for the projects themselves to the transmission required to connect them to the grid, to the interconnection queues that the wholesale power markets operate. So, issues all well known to your listeners and regularly discussed on this podcast that we'll actually need policy work to address.
David Roberts
And I wanted to ask about those, because those are, I think it's sort of the consensus these days. Those are the problems now. Like the money part is, IRA takes care of the money part, and those are now the big problems. The non-cost barriers, NIMBYism, permitting, the whole bureaucratic tangle, the lack of transmission. How do you model those? Do you try to work those effects into your models? And if so, how? To me, that's just a total black box. I would have no idea how to predict how those things go.
Trevor Houser
Yeah, there's lots of strategies you can take. It's challenging to do, and you try to tackle it through scenario analysis rather than having a specific prediction about how quickly the residents of coastal Massachusetts will become comfortable with offshore wind turbines in their sight line. So, this suite of issues definitely hits wind harder than it hits solar and storage. It impacts all of clean electricity. But wind really struggles because wind is much more geographically dependent, and where the best wind is is usually not where electricity load is. And so getting the wind from where it blows to where the power is needed requires just much more enabling infrastructure than for solar and storage. Though solar and storage do face some of the same issues.
David Roberts
Yeah. So, the story in electricity overall is that solar, utility-scale solar and storage, more or less kept up with your model predictions, right? It was mainly wind falling short that led the whole electricity sector to fall short, is that right?
Trevor Houser
Yeah. So, that actually varied across the models. Some of the Rhodium's models were pretty close for wind in part just because the timelines of the wind projects in the pipeline were relatively well known. But we fell short on solar and storage. A lot of projects slipped in solar and storage into 2024, and I think that was a combination of interest rate, supply chain, interconnection queue issues. So the projects are still in the pipeline. They just got delayed into 2024. But it is true that if you look at by 2030, where the most growth needs to happen, it's in wind.
Across all the models, wind needs to recover from its current six gigawatts, a paltry six gigawatts that were installed last year. We need to be seeing like 30 to 50 gigawatts a year of wind by 2030 being added to the grid.
David Roberts
So, it's not like solar and storage are just going to jam on the accelerator and make that up.
Trevor Houser
It's hard to imagine that if you think about, depending on the model, by the end of the decade, we need to be adding 100 gigawatts a year of clean up from the 30 to 40 gigawatts a year. Now we need to grow to that level. You can definitely see a pathway where solar is doing 50 to 60 of that, right? From the kind of 20 to 30 to 40 we're seeing now, that growth seems pretty reasonable. I think the bigger question markets can wind get from the six gigawatts a year we saw last year, up to 30 to 50 gigawatts a year by 2030.
David Roberts
From your perspective, and this is a totally subjective judgment, but speaking of the flashing yellow light, the positive story would be this is mostly about those transitory factors, and things are going to get on track again shortly. The negative story would be these non-cost barriers are deep and pathological and are going to bite harder than people thought, and there's no resolution on the horizon. So we're going to have to downgrade our models of what wind can do. Where is your subjective sense of alarm about this? On the scale of alarm?
Trevor Houser
Yeah, I mean, my guess is, just because of project timelines, my guess is that the 2030 forecast for wind is going to start coming down in the next year or two across modeling groups, just because the time it takes to build transmission, the time it takes to get wind projects going, the farther you push out that pipeline, the harder it is to meet that mark in 2030. Whereas solar, again, because of less geographic dependency, shorter project lead times, it's easier for that to make up the gap. And I think that solar will fill in some of the space if wind is slower to come online in meeting the clean electricity needs of utilities and large CNI customers. And we're already seeing that happen.
But we really do need a balanced portfolio of wind and solar. They're complementary resources and different geographic parts of the country. And we really do need to have both of those in the mix for a low-cost transition.
David Roberts
So if we're going to worry about something, we're going to worry about wind. You also mentioned that heat pumps, everybody's favorite, and by everybody I mean my favorite technology, are short of your projections. Why? Is that also about interest rates? Or is there something else going on there?
Trevor Houser
Yeah, that one's just interest rates. So here it's not so much about heat pumps specifically. It's that, because interest rates are high overall, residential construction activity has dropped off quite sharply. So, heat pumps continue to gain market share in 2023. They continue to gain market share relative to traditional AC and gas furnaces. But overall heat pump installations decline quite a bit just because construction activity is down from high interest rates.
David Roberts
So we're not going to worry about that. We think that's transitory. We think that'll pass with interest rates.
Trevor Houser
The overall market decline will, you know, I think, there's still, I would say the incentives for heat pumps relative to the cost of a heat pump installation are much less generous than the incentives for other technologies. So the incentive wedge that the IRA provided for heat pumps is pretty modest to start with. And so I think the effect of those incentives is going to be much more modest than what we're seeing for EVs or clean electricity.
David Roberts
Interesting. I wanted to ask also about e-bikes, whether you, speaking of my pet technologies, did you get down to that granularity?
Trevor Houser
We didn't, sadly, but we'll put that on the list for next year. Dave, just for you.
David Roberts
You can't just do cars, Trevor. It's all about density and biking.
Trevor Houser
I know. We'll do e-bikes.
David Roberts
Okay, so two big success stories. Well, EVs, I guess, are a success story, but two big sort of, I think, kind of eye-popping success stories are clean energy, manufacturing and emerging technologies. Let's start with manufacturing. This, it seems to me, and tell me if you agree, is kind of the headline story about IRA, just from a sort of ordinary citizen's point of view. It has sparked an extraordinary investment in manufacturing facilities, mostly, as far as I can tell, batteries, is that right?
Trevor Houser
Yeah. EV supply chain. So everything from critical minerals to battery component materials to battery cell and modules to then vehicle assembly. So you have the EV supply chain, but also in solar, which is a sector that I think a lot of people had lost hope would ever be manufactured in the US. And we're seeing pretty substantial new manufacturing investments in the solar supply chain as well.
David Roberts
Interesting. And we're seeing more manufacturing investment than modeled or is just keeping up.
Trevor Houser
None of the groups modeled manufacturing investment, actually. But I think it has exceeded analyst expectations. It certainly exceeded our expectations on the pace. And when you look at the economics of these manufacturing facilities, it is clear that the IRA is playing a pretty transformative effect. And you see it in company filings when they're talking about to their investors the role that the IRA is playing in driving their investment decisions.
David Roberts
And emerging technologies, these are sort of your capital-O, "Other technologies," gap fillers. So in that we're talking about like hydrogen, sustainable airplane fuels, things like that, what counts as emerging technologies and what do we see in that sector?
Trevor Houser
Yeah, so the three that we're currently capturing in the Clean Investment Monitor are sustainable aviation fuels, clean hydrogen, and then carbon management, which is both like direct air capture as well as point source carbon capture. Point source carbon capture is, while it was originally envisioned as a power sector play, if you remember back to like FutureGen and this idea that we were going to have a bunch of coal plants with carbon capture —
David Roberts
I don't think that dream is dead yet. I mean, the technology is dead, but I don't think the dream is dead.
Trevor Houser
Yeah. One of the many things for which cheap gas completely eroded the commercial logic was a coal-fired power plant with carbon capture. But where we're seeing a lot of it is in the industrial sector, in ethanol refineries, in petrochemical facilities that have relatively pure streams of CO2 that can be reasonably affordably captured.
David Roberts
And that's 45Q? Can't keep my 45s straight. 45Q is big, right? I mean, it's a big incentive to do that.
Trevor Houser
For these emerging climate tax, the incentive is very large relative to the cost of technology. So whether you're talking about SAF or clean hydrogen or carbon capture, the incentive is very large relative to the cost. The total federal spending right now is very small because these technologies are in their early stages.
David Roberts
Yeah, very small base.
Trevor Houser
Right. The magnitude, yeah, small base. But the magnitude of the incentive, like for manufacturing, is quite large, pretty transformative. It's very hard to imagine a lot of these projects going forward without that federal incentive there.
David Roberts
Right. And that, again, was what they set out to do.
Trevor Houser
That was the plan.
David Roberts
Attempted to do that, and it is working. It is pulling investment into those new technologies.
Trevor Houser
And I should say here for the emerging climate technologies, this is a place where the infrastructure bill actually plays a relatively large role compared to the IRA. I mean, the IRA does as well, but there were hubs programs and other incentives in the Bipartisan Infrastructure Law specifically focused at direct air capture, clean hydrogen. And so those have played a pretty important role as well.
David Roberts
It will be so interesting to see how 45V, the incentives for clean hydrogen, end up looking and end up doing. Speaking of hard to predict and hard to model, depending on how those subsidies play out, it could really radically affect where and how much clean hydrogen gets going.
Trevor Houser
It could, yeah. To date, so, we've seen a lot of green hydrogen, so electrolytic hydrogen announcements, but most of the actual investment and then the projects that have proceeded in 2023 to investment have actually been blue hydrogen projects because they are able to claim that 45Q credit, which is an existing credit, was well understood, whereas the 45V credit we didn't get final guidance on until late last year.
David Roberts
Is blue hydrogen going to be able to double up and get 45Q and 45V?
Trevor Houser
You got to choose. You got to pick one. But you have a pathway, particularly where natural gas prices are right now. You have a pathway to potential commercial viability just with 45Q for a lot of these facilities.
David Roberts
So, that's sort of the technology sectors. And I guess if we're kind of summarizing, it sounds like the overall message is we're more or less on track with kind of the big footnote being the shortfall of wind, but threatening getting to — was it 40? Is that what the model said? 40% US GHD reductions by 2030, is that right? That's the model prediction.
Trevor Houser
It's a range, like, we had, Rhodium is 29% to 42% was our range. But the midpoint in most of the models was like 39, 40%.
David Roberts
Right. And so we're more or less on track there, except for wind. Is that kind of —
Trevor Houser
And heat pumps. And building electrification is really lagging now. We don't know yet what that will look like once interest rates correct. So it may be that the trend for building electrification will be just fine once interest rates normalize, but as of right now, we're not making enough progress there.
David Roberts
Right. So, yeah, just, I guess, yell about wind one more time or yell about transmission one more time. Yell about NIMBYism one more time. Like everything, our hopes for IRA are coming down to that stuff. We can't get around it much longer. I had a couple of other sort of quasi-political questions to wrap up with. Is it true, I think that the sort of story that is going around now is that the bulk of investment, both public and private, is going into red states, is that true?
Trevor Houser
We haven't actually crunched those numbers, specifically. The Center for American Progress uses the data from the Clean Investment Monitor, and I think they have a breakdown by congressional district there. So you could look it up on that. But I can certainly, just eyeballing the league tables of state investment, particularly if you look at it relative to the size of the economy. Right.
So where is clean investment playing the largest role in the —
David Roberts
Most transformative?
Trevor Houser
Right. If you're looking at total investment, total clean investment as a share of state GDP. The top five is Nevada, South Carolina, Arizona, Tennessee, and Montana, followed by Kentucky, Wyoming, Georgia, and Kansas. Then you get to California at number ten.
David Roberts
Yeah. So in addition to all the other big question marks around IRA, to me, this is one of the most interesting and something that, sadly, one cannot model, which is what political effects this will have. You know, one of the sort of hot discussions among climate people right now is how much of this would survive if Trump took office? How much are red states attached to this flood of new investment? I suppose you probably don't have any more educated guess about that than anybody else, though.
Trevor Houser
I don't. I mean, I think that it will be a tension between potential desire by Republican leadership to roll back a marquee piece of legislation by the Biden administration in tension with the home state interests of a number of Republican members who have executives and employees calling their senator representative, saying, like, "this new factory is happening in my hometown because of these incentives, and if they go away, we won't see continued investment or we might see a factory closure."
David Roberts
Yeah, well, I hope we don't have to run that experiment.
Trevor Houser
But I leave it to you, Dave, to pontificate on whether a desire to own the libs beats out home state economic interest or not.
David Roberts
I will always bet on owning the libs, but maybe we'll find out. Another quasi-political question, and I know you are also a bit of a China expert in your spare time, and I'm curious if you have any general thoughts about how IRA and BIL — and whatever the third one is that I can never think of — how they affected other countries strategies, the passage of the bill and the subsequent investment boom, how they've affected other countries, like how China, for instance, is responding? Do you think it's galvanizing other countries? Because from Europe, you hear a lot of complaints about the sort of domestic content requirements and those kind of things being basically a form of protectionism, like we're sort of illegally subsidizing our own manufacturers, which is supposed to be against World Bank rules.
You hear a lot of that from Europe, but then you also hear from like Australia, like, they're just inspired by, galvanized by the IRA and rushing to do their own thing. So I wonder if you have any, know, China specifically, but in general of kind of the net effect on other countries aspirations and policies.
Trevor Houser
So, I guess I would tackle that in two ways. One is, what is the overall effect of the passage of the IRA? And then a second is, what is the effect of the manufacturing incentives in particular within the IRA? Right. The more industrial policy component of the IRA. So, I would say the effect of the passage of the IRA as a whole is unequivocally supportive of clean energy deployment policy in the rest of the world because it is vital for the US to meet its NDC and it is the backbone of any credible story the US can tell about how it might have a pathway to meet its climate commitment. So, since we operate in this "pledge and review" international climate architecture, it is vital that large emitting countries like the US are able to demonstrate a credible pathway to meet their targets to ensure there's incentive for other countries to do the same. And the IRA has definitely had that effect.
If you look specifically at the manufacturing incentives in the IRA, which as we talked about, have been fairly transformative in terms of the pace of manufacturing investment, I think interestingly, to date the causal chain, I think China had a significant impact on the adoption of those incentives in the US. Right. I think the build-out of a domestic clean energy manufacturing base in China, which has been primarily to serve Chinese domestic demand to date. Right.
That's where the vast majority of Chinese solar panels and Chinese EVs go, is within China. That increases the odds of the US seeking to support the growth of a similar clean energy manufacturing base in the US. I think that directionally, the passage of the IRA has increased calls in Europe for incentives for clean energy manufacturing in Europe as well. That's running a bit into some pretty tough budget realities that Europe's grappling with right now. So it's not clear how much effect that'll have. More broadly, what we're seeing right now is there's kind of two different ways to do clean energy industrial policy and to try to support a domestic industry.
And I think it's clear that in all major economies, the adoption of sustained, ambitious climate policy requires having a domestic industrial base that's getting to benefit from that, just for the politics of it, right?
David Roberts
Just politically, you mean?
Trevor Houser
Right. So put aside, whether you believe, just as an economic matter, industrial policy as good or bad, on which economists will debate until the end of time, is a necessary ingredient in the formula for political success on climate, it's pretty important. That industrial policy can take one of two forms. One is incentives and subsidies, lowering the cost of clean energy manufacturing. I see that as broadly supportive of the pace of global deployment. It expands supportive of our climate goals. Right. There's no tension there, really, between industrial policy and solving climate change. Those incentives, like the ones we see in the IRA, are expanding the amount of production capacity available to the world and lowering the cost of producing clean energy technology.
Right? And like we talked about in the case of utility-scale, wind and solar, we need to build a lot of this stuff. We need a lot of manufacturing capacity. Now, your Econ 101 professor would quibble that it is distorting optimal allocation of capital in the market by doing that.
David Roberts
Yes, which is just to say the intuitive Econ 101 take here is if you want to decarbonize as fast as possible, you buy the cheapest possible decarbonization technology from wherever it's made. That's how you maximize speed. And anything else, almost definitionally reduces you from your optimal speed.
Trevor Houser
But actually, the subsidies don't cut against that. The subsidies make things cheaper, right? So they increase speed. So subsidizing expanded manufacturing capacity makes clean energy technology cheaper globally. It does not raise the cost of it for anyone. The Econ 101 professor's quibble would be that it's allocating too much capital to clean energy investment. That should be going somewhere else from like an economically optimal standpoint, right? I'm okay if at the end of the day, in solving climate, the cost was some slightly suboptimal global capital allocation, that seems like not too bad of a deal. So if we're doing industrial policy through incentives and subsidies, then we're making clean energy cheaper, we're accelerating the pace of deployment, broadly defined.
Right. The other way to do clean energy industrial policy and to support a domestic industry is through barriers and tariffs, where you are not allowing in goods from other countries.
David Roberts
They are the Econ 101 objection by —
Trevor Houser
Correct. And that is where there actually is a tension between the pace of clean energy deployment and your domestic industrial policy goals. Right? To date, the US and most other countries have actually been primarily leaning on the former: let's support the expansion of a domestic clean energy industrial base through subsidies and incentives.
David Roberts
A couple of solar tariffs.
Trevor Houser
Right, generally. And then there have been a couple of very high profile and important cases of tariff barriers. I think one of the defining challenges for the next decade for policymakers in the US and around the world is as these incentives build out a meaningfully sized domestic industrial base, those domestic industries are going to start pressing for tariff protection. And this is where it's going to require pretty clear-eyed and prudent policymaking on how to both support the growth of domestic industrial capacity and ensure that we're delivering clean energy at scale and as cheaply as possible around the world to meet our climate goals.
But that's to date. Fortunately, the industrial policy goals of the US and China and the European Union and the climate goals have actually not been that much in tension.
David Roberts
Right. Perhaps worth mentioning that Trump proposes a wide range of these latter strategies, tariffs and barriers, and etcetera. The Econ 101 objections do not seem to trouble him much.
Trevor Houser
They do not.
David Roberts
Okay, so wrapping up, I've actually got a whole pod on this subject coming a little bit later, but I'd love to have just a few thoughts from you based on your close-up view of how IRA is put together and how it's performed. What, in the unlikely event that Biden won and Dems won Congress again, both houses of Congress again, in the highly unlikely event that Dems got another trifecta in 2024, what would you advise for IRA part two, if you had a couple of top items?
Trevor Houser
So,if we broaden that to legislative, legislative agenda too. So the first piece would be for the technologies that IRA made cost competitive: what are the non-cost barriers standing in the way to the speed and scale we need? Right. So that's one category. And then the other is —
David Roberts
Can't do that through reconciliation.
Trevor Houser
Cannot do that through rec — well, there's some pieces of it you might be able to, but most of it you can't do through reconciliation. The second is the sectors where we're just now starting to build out the technologies. And like we talked about with emerging climate technology, IRA provides important early technological development incentives. But for things like industry and agriculture and land use, as transportation, surface transportation and power decline, the largest source of emissions in the US will be industry.
David Roberts
It was just last year, or the year before, that transportation overtook electricity.
Trevor Houser
2016. Transport overtook electricity.
David Roberts
Was that 2016?! What is time even mean?
Trevor Houser
Time's a circle.
David Roberts
Time is a flat circle. Okay, so cars overtook electricity in 2016 as the top emitting sector in the US. Presumably, the IRA is going to bend that curve down. They're both going to go down. And at some point, the industry is going to become the top sector emitting in the US. What year does that happen?
Trevor Houser
All right. Well, if you count the industry like we do, which includes the process emissions and the production and transport of oil and gas — so important caveat there. If you include that, then by 2026, the industry will be the largest source of emissions in the US.
David Roberts
Real soon. So electricity is going down. Is transportation going down yet?
Trevor Houser
It is, yeah. Gradually. More gradually than power, but it's going down.
David Roberts
And is the industry still going up?
Trevor Houser
The industry is still going up.
David Roberts
All right, 2026 it is. When is that going to happen? Anyway, I interrupted your second half of your answer. Your first half is the non-cost barriers to the scale technologies which are stressing everybody out right now. Part two is what again?
Trevor Houser
Is the harder-to-decarbonize parts of the economy, which account for about half of the emission. So that's the industry, which we've pretty much not touched at all. From a decarbonization standpoint.
David Roberts
Yeah, that's interesting.
Trevor Houser
Agriculture and waste and hard to electrify transportation. So sustainable aviation fuel, maritime transport. Now, the IRA is making some important early technology development investments, but scaling those is going to require a different set of policies.
David Roberts
Yeah. Well, on that note, one of the things I'm trying to wrap my head around is we've got this set of tools that we used for this first round of legislation, mostly, the IRA, that seem to be working as projected as planned. Are those tools, is it just a matter of taking those same tools and applying them to these new sectors, or are these remaining sectors going to require novel policy instruments?
Trevor Houser
I think some of the same tools could work. So if you were thinking about cement as a segment and what do we do to produce low carbon cement, you could certainly imagine a tax expenditure based approach where it is a tax credit for production of low carbon cement. That would certainly work. There are other policy instruments that can work as well. You can develop portfolio standards for industrial goods. You can use Clean Air Act regulations for industrial sources. There's other options as well. Tax credits have proven to be the most popular and durable form of climate policy in the US.
David Roberts
For better or worse.
Trevor Houser
Yeah. The appetite in the future to continue using that as the primary vehicle for federal incentives, I think will depend in part on what the budget looks like and appetite for more tax-based spending.
David Roberts
Yeah, agriculture, though, it's a whole different pod, I guess, but it's somewhat of a mystery to me. I guess I just kind of want to emphasize that it's not obvious. If you sat down to write IRA two today, what you would write.
Trevor Houser
Yes, that's correct.
David Roberts
I don't know that — we're not nearly as prepared for these next sectors as we were for what we did this time.
Trevor Houser
Yeah, that's absolutely right.
David Roberts
All right, well, this is super interesting, and presumably you guys are still tracking — I guess one thing to say by way of wrapping up, and it's just an important caveat, is this is all relatively new. These are huge pieces of legislation, transformative. There are going to be long-term effects too. And as we said, lots of the money still hasn't gone out the door yet. Some of the tax credits are not even in place yet. Some of the other programs are not stood up yet. So should just flag like these are early days.
Trevor Houser
Absolutely. And we'll be on the Clean Investment Monitor. We'll be every quarter releasing updated investment numbers and updated public spending numbers. And users can download the full underlying database to do their own analysis on as well.
David Roberts
Fun, fun, fun. All right. All right, Trevor, thanks so much. It was great to talk to you.
Trevor Houser
Yeah, thanks, Dave.
David Roberts
Thank you for listening to the Volts podcast. It is ad-free, powered entirely by listeners like you. If you value conversations like this, please consider becoming a paid Volts subscriber at volts.wtf. Yes, that's volts.wtf. So that I can continue doing this work. Thank you so much and I'll see you next time.
How's IRA doing?