Washington lawmakers have spent trillions of dollars fighting poverty based on a four-decade old measure that most economists agree is wrong.

Many national efforts to help lower-income Americans rely on the government's official poverty line, developed in 1965. But that dated statistic includes only people who received cash-based assistance from the federal government and doesn't take into account things like food stamps, tax credits and free school lunches.

"The official measure is pretty much worthless," said Ron Haskins, a senior fellow at the Brookings Institution, a think tank, and former senior adviser on welfare policy for President George W. Bush.

The problematic measure undermines anti-poverty programs by making them look less successful, economists say. It also distorts today's poverty landscape by misconstruing many key statistics. That can lead Congress to spend money on the wrong kinds of programs and advocates to push for the wrong kinds of initiatives.

The fact the official poverty measure has stuck around so long despite its well-known problems shows just how little attention Washington pays to the issue.

But that's starting to change.

In 2011, the U.S. Census Bureau issued new poverty estimates based on a model that fixes some of the official measure's basic shortcomings, such as skipping over non-cash assistance.

Economists say the supplemental poverty measure paints a more accurate - and bleaker - picture of America's bottom tier.

According to the government's official poverty measure, 15 percent of Americans, or 47 million people, were in poverty in 2012, the most recent data available. That's not much different than the poverty rate when the government started measuring, which has stayed between 12 and 15 percent since the 1970s.

But by using the new formula, 49.7 million Americans were in poverty in 2012, a higher number than the official measure. In addition, it showed that 12.9 percent of Americans over 65 are in poverty, compared to just 8.4 percent by the official measure, while less children are in poverty than previously thought.

Its findings undercut policymakers' major arguments on both sides.

The supplemental measure finds government programs like food stamps have actually cut poverty in half since 1967, undermining a popular conservative argument that government welfare programs keep people in poverty.

"The safety net programs have been much more successful than people give them credit for," said Isabel Sawhill, a senior fellow at Brookings who co-directs the think tank's Center on Children and Families with Haskins. "In the war on poverty, we won."

The new measure also indicates that fewer children are in poverty than the official measure indicates, an inconvenient statistic for liberal advocates calling for more federal spending to solve the problem.

"If you want to spend more money, it's good to have the highest poverty rate you can," Haskins said.

But this new way of measuring poverty isn't perfect. The supplemental measure doesn't take into account benefits from massive government health care programs like Medicaid and Medicare.

And because the supplemental measure is so new, there are no trends to compare it with, which is one reason the official measure is still around.

"If you just knew the poverty rate for this year and last year and you didn't know anything else, what would you be able to say?" Sawhill said.

There is a team of economists trying to rewind the supplemental measure back five decades to more accurately estimate historical trends of poverty, but their work is still in progress.

There are other options economists are considering.

One theory looks not at what people are receiving from the government but at what they are buying themselves.

The "microwave theory," as it's colloquially known, tallies up everything Americans buy to raise their standard of living, from appliances like dishwashers and microwaves, to electronics like laptops and printers.

The measure is championed by conservative economists such as Aparna Mathur with the American Enterprise Institute, who says consumption is a better indicator of people's standard of living than income.

"It really tells you much more about how well off they are today relative to, say, 20 years back," she said.

By the measure of what Americans can afford to buy, inequality has only grown by a small margin over the decades, so fears of inequality disparities are overblown, Mathur testified in a Jan. 16 Joint Economic Committee congressional hearing.

"The typical low income household possesses many more appliances and gadgets that have traditionally been considered the preserve of the rich, than at any time in history," her testimony read.

The consumption measure is a great idea in theory, but the real truth to measuring poverty is that there's no perfect way to do it, Haskins said.

That's because poverty is multi-faceted, economists say.

David Grusky, the director of the Stanford Center on Poverty and Inequality, suggests we start measuring poverty the way we measure unemployment in America -- by a whole host of indicators.

"The more measures the better," he said. "I don't think we should settle on a few."

Sawhill agrees. There are many different reasons why people are poor, and many variations on how long they stay poor, so one measure just won't cover it all, she said.

"By digging down a little bit, you could begin to have a different kind of debate," she said, adding, "Policy has to be tailored to the reason why people are poor, not just the fact they are poor."

Grusky argues that even with new measures of poverty, that hasn't happened yet. Instead, policymakers have continued to focus on the symptoms of poverty instead of its causes.

He draws on what he calls the log-factory metaphor: A town downstream of a pollution-prone log factory can employ social workers, doctors and government officials to treat the illnesses and missed productivity and dozens of other social problems caused by pollution. Or it can bypass all that trouble by going directly to the source and regulating the log factory.

The causes of poverty are like the log factory, he argues: They could be nearly eliminated by going to their source.

"There are complicated effects of poverty, but not causes," he said.

Solving poverty requires investment in both education and creating more jobs, both conservative and liberal economists agree.

And Grusky said that is where policymakers fall short: failing to make that investment.

He, among others, theorizes that investment in fighting poverty's causes costs far less than tackling its downstream effects.

"It's just a matter of saying, 'Actually, if we thought seriously about it and really fought a second war on poverty as if we meant it, because we know the costs are so dire, we could do it,'" Grusky said.

But first, policymakers need to agree on how to measure poverty in the first place.