Everyone knows that in math, 4-2 = 2. But that's not always true when it comes to tax policy. Sometimes (often, in fact) 4-2 = 3.

In other words, if we cut certain tax rates in half, we won't necessarily cut tax revenues in half. Sometimes they'll drop by far less.

How so? Because when the government cuts confiscatory tax rates, people tend to work more, save more and invest more. That improves the economy, which means tax collections can also increase. Lower tax rates also improve compliance since people have less incentive to avoid and evade the IRS.

Problem is, Congress doesn't count these "supply-side" benefits of lower tax rates. Instead, before lawmakers vote on a spending or taxation bill, they give it to the Joint Committee on Taxation. The JCT is supposed to "score" it, or explain what the revenue effects of the legislation will be.

If the bill calls for a tax cut of, say, $100 billion over 10 years, the JCT generally subtracts $100 billion from its revenue predictions. But this approach simply doesn't work, because it ignores the fact that certain types of tax cuts -- such as lower tax rates -- tend to boost the entire economy.

Remember the 2004 American Jobs Creation Act? When lawmakers were debating that bill, the JCT guessed the entire tax cut package would cost some $4.5 billion in its first year. That was off by just a bit. In fact, the change brought in an additional $16 billion in the first year.

That's because the JCT underestimated a provision of the new law that substantially reduced the second layer of tax on companies bringing foreign earnings back into the United States. As that cash flowed home in response to the lower tax rate, the country enjoyed a $350 billion investment that in turn generated a big boost in tax collections.

Unfortunately, the JCT's 2004 error wasn't its first. It also wildly overestimated the "cost" to the government of tax cuts in 1997 and 2003 (JCT's "static scoring" on the 2001 tax bill was okay since that legislation focused on the types of tax cuts -- such as rebates and credits -- that don't have any impact on growth).

It's clear we need a better analytical approach, one that accounts for the fact that cuts in tax rates (especially on investments and savings) can boost the economy. Washington insiders call this "dynamic scoring," but average Americans would call it, simply, reasonable.

They know that if a family is allowed to keep more of any additional income (because the government cuts the family's tax rate), the family will save some -- and spend some -- of the additional dollars. A family's savings especially help boost economic activity: firms that want to expand their operations have a greater pool of funds to draw from, and lower taxes on savings means that interest rates will likely fall, thus enabling lenders to back more entrepreneurs, the real drivers of the U.S. economy.

That's the dynamic effect economists love to see in tax policy change. As a bonus, individual families make better choices about how to spend their money than Washington does, so when they're allowed to keep more, the country will automatically benefit.

Luckily, the government is finally moving in the right direction. This year, President Bush proposed creating a Dynamic Analysis Division within the Treasury Department's Office of Tax Analysis. This group would advise the President and other key policymakers on how proposed changes to U.S. tax policy affect economic activity. In other words, it sets the stage for "dynamic scoring" of tax bills.

This simple step should help us draft more intelligent tax bills. That's critical, because all tax cuts are not created equal. We've already seen that the 2003 and 2004 cuts that slashed rates on productive activity boosted growth and increased tax revenues, just as the dynamic revenue models at The Heritage Foundation had anticipated they would. Sensibly, Congress recently extended those cuts through 2010.

But we also need to avoid ill-advised tax cuts, such as the tax rebate of 2001.

Lawmakers enacted that because they assumed it would "put money in the hands of consumers."

But as good dynamic models predicted, the rebate didn't decrease the cost of working or investing, and those are the two major factors that drive economic activity. The windfall may have helped people save a little cash, but it didn't change the tax rate you face when deciding to put in some overtime, take a second job or start your own business. In short, credits like this one don't change the incentives to work, save and invest.

President Bush says one of his long-term goals is to fundamentally rework the tax system. That's a worthy goal. No doubt it'll be much easier to achieve the simple and fair flat tax system the country deserves when policymakers can see the real effects -- good and bad -- of their tax policy decisions.

When they understand the benefits of the "new math" of dynamic scoring, they'll understand how we can add to the economy by subtracting confiscatory tax rates.

Edwin Feulner is president of The Heritage Foundation (heritage.org), a Washington-based public policy research institute and co-author of the new book Getting America Right.