The IRA is projected to reduce emissions by 40%. Not everything is so simple.

The question is how much.

The most popular figure is the one repeated by the president, scientists, and journalists — 40 percent. In a statement released shortly after the deal was reached, Democratic senators Charles E. Schumer (N-New York) and Joe Manchin III (W-Va) said the new bill would cut emissions by 40 percent from 2005 levels by 2030. This figure was later confirmed by the results of three independent modeling groups. The Rhodium Group, an economic and energy research firm, estimated that the bill would reduce emissions by 31 to 44 percent by 2030; Climate think tank Energy Innovation predicted a 37 to 41 percent cut; and a group of Princeton University researchers called the REPEAT project estimated a carbon reduction of about 42 percent.

The agreement between the senators’ claims and the projections isn’t a surprise — modeling teams advised Capitol Hill staff about the likely effects of the deal before it was made public, said Jesse Jenkins, co-director of the REPEAT modeling project at Princeton.

That 40 percent figure would be repeated at international climate talks and in presidential speeches for years to come. It marks progress toward the president’s top climate goal of halving emissions by 2030 — and could offer some hope to the millions of young people who have been drawn to climate action in recent years.

But is it correct? It depends on how you measure and what you measure with.

At the heart of these predictions are scientists’ very sophisticated models of how the economy works, including how energy is used, which can provide useful predictions about the future and always several is not accurate. As one popular modeling saying goes, “All models are wrong; some are useful.”

The energy models used by Rhodium, Energy Innovation and Princeton researchers are complex systems of equations, spreadsheets and data that attempt to represent all the energy used in the United States over a period of time. These models allow us to estimate how many solar farms will be built after tax credits are introduced to make them cheaper, or how many Americans will buy electric cars in the next 10 years.

The fact that all three independent modeling groups produced similar findings is a good sign for the results. But there is still reason to believe that the reality may differ from what models suggest the bill’s impact will be — or what the public might expect.

On the one hand, models predicting a 40 percent drop in carbon emissions may be overly optimistic. Jenkins, a professor of mechanical engineering at Princeton, says one of the main challenges is predicting how quickly consumers, utilities and businesses will switch to clean technologies.

“The most important thing about our model, which is an abstraction of the real world, is the assumption that financial considerations drive decision making,” he explained. The models assume that people are rational actors who make their decisions based on costs and benefits; in the real world, this is not always the case.

This means that if it is cheaper to build a wind farm than a natural gas plant, or it is cheaper to buy an electric car than a gas car, the model predicts that more wind farms will be built and more electric cars will be purchased. The REPEAT model currently predicts that all cars sold in 2030 will be electric cars, as electric cars are projected to be cheaper than gas-powered cars by then. But in the real world, some consumers will be afraid to switch to electric vehicles, even if they are cheaper, simply because they don’t see enough car chargers in their area.

Similarly, wind farms and solar panels may be stymied by local residents who find them unsightly. Long-distance transmission lines, which would be needed to transfer renewable electricity from one state to another, could also be stymied by red tape.

External economic factors may also slow the shift away from fossil fuels. Ben King, associate director of climate and energy at the Rhodium Group and one of the authors of the IRA analysis, says cheap fossil fuel prices and faster-than-expected growth could lead to a slower-than-anticipated transition to clean energy.

2) There is potential for growth

There is also reason to think that 40 percent is not underestimate for the effects of the bill. Jenkins notes that neither model can effectively predict technological advances spurred by public money — such as research and development funding that causes the costs of solar panels, wind, carbon capture and storage, or batteries to plummet. These cost changes, he argues, could make the transition to clean energy even faster than expected, but are difficult to predict in an energy model.

The models also do not attempt to predict changes in state and federal policy. But over the next few years, many states and cities are likely to adopt new climate policies, such as requiring electricity to come from renewable sources or phasing out gas-powered cars.

“This bill makes it cheaper for any other jurisdiction in the country to raise its ambitions and policies,” Jenkins said.

It would also make it easier for the Biden administration to impose tighter limits on emissions from cars and power plants — which, in turn, could further reduce CO2 emissions.

However, there is no “sure thing” about modeling. The reduction in emissions from the IRA may be higher or lower than the 40 percent estimate; at the moment, modelers can only make their best guesses about how the future will develop. But, Jenkins argues, the result isn’t much different from estimates of the bill’s cost.

Because clean energy IRA provisions consist primarily of tax credits, it is difficult to predict how many of these credits will ultimately be claimed and how much the bill will cost the government and taxpayers.

“Forty percent is an imperfect estimate,” Jenkins said. “But I think that’s a pretty good estimate.”

3) Progress is already baked

Depending on how you read that 40 percent estimate, it can be misleading. This is a case where the models may be correct but not widely appreciated. The bill is expected to cut emissions by 40 percent in comparison with 2005 levels — not compared to current US emissions. That’s because emissions have already dropped significantly since 2005. Between 2005 and 2020, CO2 emissions fell by about 21 percent, largely due to the shift from highly polluting coal to less polluting natural gas. (The COVID-19 pandemic also caused a sharp drop in emissions, as millions of cars and planes ground to a halt virtually overnight.)

Emissions are expected to slowly decline over the next eight years thanks to cheap solar and wind power and a gradual shift to electric vehicles.

Indeed, according to the same three modeling groups, emissions are expected to decline by 24 to 32 percent by 2030 — even without the Inflation Reduction Act.

That doesn’t mean, of course, that the bill won’t have an impact. While this may not be as dramatic a shift as it first seems, in a world where every extra ton of CO2 that is not emitted into the atmosphere can help curb global warming, an additional 10-15 percent reduction in emissions will help prevent serious environmental damage.