From Capitol Hill to the Motor City, marble statehouses to mud-strewn shale plays, lawmakers, policy wonks and industry leaders repeatedly invoke U.S. government energy forecasts to call for more drilling, more fracking and more nuclear plants.

These predictions are made by the Energy Information Administration. They are ostensibly apolitical, certified with a federal seal of approval and are ripe fodder for the journalists, analysts and government agencies responsible for painting a picture of the country’s energy future.

But in truth, some experts say, we’d have better luck calling Miss Cleo.

“The idea that these things they purport – that they say they’re scenarios, not projections – this is intensely disingenuous,” says Travis Bradford, director of the energy and environment concentration at Columbia University’s School of International and Public Affairs. “People are making substantial capital investment decisions based on what they perceive to be the perceptions and the expertise of the EIA.”

EIA models explore virtually every conceivable aspect of the energy industry: how the U.S. will generate its electricity, how Americans will heat their homes and fuel their cars, whether coal and oil will still be viable 50 years from now, and how fast solar and wind power may expand. The agency calls the analyses “outlooks” and “scenarios,” and the results inform decisions at every level of government; last year, they were the foundation of the State Department’s analysis of the controversial Keystone XL pipeline.

As a central clearinghouse for all things energy, “EIA has been a tremendous national resource,” says Katherine Hamilton, a co-founder of the energy consulting firm 38 North Solutions.

But plummeting global oil prices, surging renewable power and U.S. commitments to slash heat-trapping carbon emissions have made America’s energy and electricity sectors ever more volatile. And in an age of global warming, when it is perhaps needed the most, the EIA – granted governmental independence by Congress and burdened by an expectation that it remain policy-neutral – has largely failed to adapt, experts say.

This isn’t precisely true. Congress, in establishing the EIA in 1977, used language stating “that EIA data and analysis is not subject to approval by the other parts of the executive branch as a way to keep EIA independent from political influence,” agency spokesman Jonathan Cogan says. So it’s largely “principle” or “practice” that keeps the EIA policy-neutral, Cogan says, not the strict letter of the law.

But the outcome is the same, critics say. Each spring, the EIA produces its Annual Energy Outlook, a roughly 150-page rundown on the future of America’s sprawling energy sector. Every prediction comes with a range of possible scenarios but is anchored by a “reference” case: a graph showing what would happen under current policy or “business as usual.”

The theory is sound: “By having a reference case under current laws and regulations, we are able to isolate what the impacts of the proposed new regulation might be compared to no change from current regulations,” Cogan says.

And, adds Richard Newell, the head of the EIA from 2009 and 2011 who now directs the Duke University Energy Initiative, using only current laws and regulations helps protect the “independence and content of EIA reports” – it helps prevent the agency from being accused of supporting or opposing potential new laws or regulations.

But the approach, experts say, is both flawed and misleading.

“They do fabulous work on a lot of things. But somehow this particular exercise is just abysmal,” Bradford says.

Take a basic graph included in the most recent Annual Energy Outlook: “Electricity generation by fuel, 1990-2040.” The outlook purports to show just how much electricity the U.S. will churn out from energy sources like natural gas, coal, nuclear power, oil and renewables such as wind and solar over the next 25 years. The model, however, ignores not merely far-off political possibilities, but current economic realities, some experts say.

The coal and nuclear sectors, for example, are shown falling slightly but otherwise remaining level; from 2015 through 2040, each respectively retains around 34 and 16 percent of the U.S. electricity market. But while coal and nuclear power are expected to remain major sources of electricity in the U.S. for decades to come – “the narrative ‘coal is on its way out’ is a bit overstated,” former Obama White House energy adviser Jason Bordoff says – EIA’s version of the future is one that “mathematically cannot happen,” Bradford maintains.

For one, utility companies that operate nuclear plants, faced with cheaper natural gas, solar and wind – plus federal age limits for the plants – may decommission as many as 15 aging nuclear plants in the decades ahead and bring only a handful of new plants online, both Bradford and the EIA say. In the coal sector, dozens of creaky generating stations are expected to be retired this year alone.

In other words, for the shares of coal and nuclear in the market to remain level, power companies would need to embark on perhaps the largest and fastest building spree in history, Bradford says. It’s a future absolutely no one is talking about – except, in its own way, the EIA.

“Projected coal-fired generation is very sensitive to future [natural] gas prices,” Cogan says. If those prices tick up – as the reference case assumes – it becomes “very attractive to run the coal plants that remain in service,” thereby allowing coal to keep hold of its market share. And when it comes to nuclear power, the spokesman adds, the future of existing plants is not at all clear, as the Nuclear Regulatory Commission may choose to extend their life spans.

Federal tax credits for renewables like wind and solar, meanwhile, are subject to the fickle nature and partisan politics of Congress.

Others outside the agency, though, dispute the math. The use of wind power alone rose by nearly 5 percent in 2014 compared with the year before, while new solar installations expanded by more than 30 percent, making solar second only to natural gas in terms of new generating capacity. This surge, critics say, casts doubt on EIA’s description that renewables’ share of the electricity market will grow an average of only about 0.1 percent a year through 2040.

“It’s affecting us in a very negative way,” says Michael Goggin, senior director of research at the American Wind Energy Association.

Indeed, those invested in maintaining the status quo – coal companies, oil companies and the lawmakers they support – could wish for nothing better: America’s authority on the future of energy says fossil fuels are here to stay in a big way. And in effect, the EIA’s approach of not taking a stance can, paradoxically, be seen as having one.

That ammunition – in the form of seemingly authoritative information from a nonpartisan federal agency – is often most lethal when used as a counternarrative.

“Nobody goes out and says, ‘Hey, EIA says we’re going to use a bunch more coal, so let’s build a bunch more coal plants,'” Bradford describes. “Instead, when people say, ‘There’s going to be a solar revolution, it’s going to grow much more dramatically than anyone’s expectations,’ ‘” [which, in fact, it has] “then inevitably someone will go to EIA, pull the reference case, and say, ‘This can’t be true.'”

He adds: “It’s an antidote to anyone predicting other behavior.”

EIA outlooks reverberate across the economy and around the globe. And while there’s no way to know just how many millions of dollars have been invested by companies based on the agency’s outlooks, smaller firms looking to make waves in the energy sector may particularly be led astray.

“When EIA gets it wrong, there are people who pay a price who don’t have analysts at their own companies,” Webber says. “We look to official government statistics to give us guidance.”

Even big corporations that can afford to keep an energy analyst on staff or fork over big bucks for an independent analysis see EIA assumptions as a crucial “part of the environment, part of the mix that companies may consider as they develop their future investment plans,” says Jim Burkhard, chief researcher for global oil markets and energy scenarios at IHS Energy, a consulting firm.

And that’s before looking at Congress or the nation’s statehouses. On Capitol Hill alone, just five months into the current legislative session, close to two dozen bills already referenced EIA statistics to justify new policies dealing with billions of dollars in subsidies, tax breaks and other incentives for different parts of the energy sector.

“EIA energy forecasts inform major policy decisions about infrastructure and energy extraction,” says Anthony Swift, a staff attorney with the Natural Resources Defense Council.

The long-term consequences ultimately add up to far more than dollars and cents. The U.S., joining other nations, has committed to reining in heat-trapping carbon emissions to avoid a rise in the global temperature of more than 2 degrees Celsius – a higher increase is widely seen as the tipping point toward catastrophic climate change.

Analyses by other agencies, such as the International Energy Agency – made up of 29 nations from Europe, Asia and North America – use calculations that assume a general commitment to that 2-degree target through the wider expansion of solar and wind power and broad declines in coal and other fossil fuels. Called the “New Policies Scenario,” the IEA makes this one of its most prominent forecasts.

The EIA does offer alternative scenarios that make some account for different energy prices, which in turn can reflect policies addressing climate change. But Swift and others say these are overshadowed by the policy-neutral reference graphs – and often ignored by the agency’s “Today in Energy” blog posts. The “business-as-usual” scenarios, in predicting no substantial drop-off in coal for decades to come, point to something far different from just 2 degrees of warming.

In fact, if the U.S. were to continue relying so heavily on fossil fuels, it would help fuel 6 degrees of warming worldwide, Swift says.

“Federal agencies tend to focus on the scenarios that EIA makes most prominent, and those are not the climate scenarios,” he says. “If we’re serious about addressing climate change – which we have to be – the scenario which we have committed to should be front and center in the projections that EIA produces, and it simply isn’t.”

Others, however, aren’t so sure.

“It’s a valuable exercise, not necessarily to predict the future, but to say, ‘Here’s what the world looks like amid current policy,'” says Bordoff, the former White House energy adviser and director of Columbia University’s Center on Global Energy Policy. “It helps you understand which policy drivers, which technology innovations will help most significantly move the needle.”

Plus, he adds, while coal may not exactly remain level, “predicting the future is always difficult.”

Hamilton, of 38 North Solutions, agrees. The EIA, she argues, has made the best of being “constrained by data and the timelines they use to produce their reports.”

Yet where some see inaccuracy – the kind inherent to any kind of prediction, whether regarding gas prices or football spreads – others see a pattern of bias, one that while not necessarily intentional, ultimately has implications that can shake broad sectors of the economy and stymie climate change efforts around the globe.

“All forecasts are going to be inaccurate. What is concerning is when you see consistent systemic bias in your projections from EIA,” Goggin, of the American Wind Energy Association, says. “There seems to be consistent bias in EIA’s projections against renewable energy, and that’s a different thing from being inaccurate.”