A pure retail sales tax without exemptions or transition relief ought to have a positive impact on growth. First, switching from an income tax to a consumption-based tax would lead to greater savings and investment. And that should increase productivity and the pace of output growth.

There’s a subtler route, too. The effective double taxation of existing capital during the transition to a national retail sales tax would generate windfall revenues and thus allow a tax-rate reduction that stimulated growth.

However, the world is not quite that simple. Many forms of saving—including pensions, 401(k) plans, and individual retirement accounts—already receive consumption tax treatment, and a significant share of corporate income is currently untaxed. Moreover, under a national retail sales tax, the likely provision of transition relief for existing assets could reduce the effect on saving further (it’s hard to imagine that sophisticated lobbies would accept double taxation without a fight).

Several analysts have constructed models capable of generating realistic estimates of how tax reform would affect growth. The most complete model, developed by David Altig and colleagues (2001), simulates the effects of moving from the current system to a flat-rate consumption tax.

Their analysis—which assumes a less generous demogrant (cash rebate) than proposed by national retail sales tax advocates, some transition relief for existing assets, and no avoidance or evasion of the new tax—finds that the economy would be 0.6 percent larger than otherwise after two years, 1.8 percent larger after 10 years, and 3.6 percent larger in the very long run. But here, as almost everywhere, the devil is in the assumptions. Plausible allowances for avoidance, evasion, and erosion of the statutory tax base for political reasons, along with a more generous demogrant, would reduce these estimates.