There was a lot going on during Wednesday night’s Republican debate: Lots of yelling and lots of numbers, and then yelling about numbers.

The figures mostly came from tax plans presented by the 10 GOP frontrunners, plans that would supposedly streamline the tax code, increase the financial health of the economy, and improve life for families. But if Wednesday night proved anything, it’s that when it comes to politics, the math of how these plans will actually work is kind of beside the point.

For example, Ben Carson’s plan for improving the tax system calls for a flat tax, and the rate he proposes is Bible-based, derived from the notion of tithing, or giving 10 percent of income. During the debate, Carson defended his plan against accusations that it would wind up costing the U.S. as much as $2 trillion due to the massive tax cuts it would provide for higher-income Americans. Instead, he says, the real rate would be closer to 15 percent, and the plan would involve eliminating deductions and loopholes. As Josh Barro noted in the New York Times, Carson’s plan is based on the idea that everyone and everything would be taxed. That would require eliminating exemptions, including those for nonprofits, low-income families, and charitable donations. Moderator Becky Quick told Carson that, according to economists she’d spoken, even a 15 percent rate would decrease revenue by $1.1 trillion. His final rebuttal: “When we put all the facts down, you'll be able to see that it works out very well.”

(Ted Cruz’s recently unveiled plan is pretty similar to Carson’s, except he’s sticking with 10 percent, and would derive additional revenue from a 16 percent business tax. He also casually mentioned that he wants to eliminate the IRS and have Americans file via postcard.)

Trump’s plan, which streamlines the seven tax brackets into three and asks those who make less than $25,000 to pay nothing and merely send the government a piece of paper saying “I win,” was also criticized as financially implausible. The main concern is that the top tax rate would be reduced to 25 percent, from current levels which hover near 40 percent, meaning a giant loss to the government’s coffers. His response to such concerns didn’t address the actual economics of the plan, but rather focused on the fact that it had found a supporter at CNBC. “Larry Kudlow, who sits on your panel, came out the other day and said, ‘I love Trump's tax plan,’” Trump yelled into his microphone.

Will any of these plans actually work? The conservative Tax Foundation analyzed the tax proposals of the candidates and found that most left fairly large holes in government revenue over the course of a decade, even for estimates that accounted for economic growth. Within a decade, they find that Cruz’s plan would allow for GDP growth of about 14 percent, with a revenue decrease of just over $750 billion. Trump’s plan doesn’t hold up as well: 11.5 percent growth and a $10 trillion revenue decrease. Rand Paul’s plan, which involves a flat tax of 14.5 percent performs better under the Tax Foundation’s scrutiny, 13 percent growth with no decrease when accounting for economic growth. But Paul’s proposal has been criticized for not actually being all that flat, and retaining many of the loopholes and deductions such a plan is supposed to weed out.

The struggle to make the numbers work shouldn’t be surprising. Deducting a significant revenue stream by cutting taxes on the richest Americans without somehow replacing it means that there will ultimately be a gap. That’s how subtraction works. Some candidates have said that they would help make up for the loss of revenue by cutting spending, but on Wednesday, the details of precisely where those cuts would happen were scant. Unless candidates were able to significantly stem and decrease spending (a process that has proved challenging in recent years) the numbers simply don’t add up to a balanced budget.

That’s a problem for Republicans, because it could mean a growing deficit, something they revile. So their economic teams are left with the task of making the numbers work. One option could be to propose concrete policy initiatives like closing loopholes and deductions, tweaking government payments, and other measures. But there’s another way to make it work: Build ridiculously rosy models of economic growth. That seems to be the path most of these candidates are taking.The theory is that by reducing taxes so significantly, they can stimulate growth in other areas which would make up for the losses. But even so, as Barro also notes, some of the math just seems outright implausible, like Kasich’s assumption that he can add about $250 billion to government revenues within 10 years despite serious tax cuts. Of course, there’s an even more cynical possibility: It could be the intention of these candidates to create a dearth of revenue, which would cripple portions of government, forcing cuts.

Even skeptics might understand the appeal of flat-tax proposals. At a time when inequality is growing, American wages are slow, and a complex tax code allows for all manner of manipulations, (especially by those who have the money to hire experts and the motive to protect high incomes), a system of taxation that places a seemingly equal burden on everyone may feel like the best course of action. But these plans run the risk over oversimplification. The current progressive tax code, with all its flaws and failures, is so complex because it is meant to serve a broad, diverse spectrum of Americans, and their needs. Unwinding the system won’t be as easy as simply levying a single tax rate.

In the end, the specific numbers, which will be tweaked and retweaked, are perhaps less important to voters than the ideas they represent. The problem is, of course, that tax plans aren’t just a set of ideas, and, eventually, revenues must come in and, somehow, bills must be paid.