While Congress passes a Republican plan to reduce taxes by $245 billion over seven years, an even more historic tax debate may be in the making. House Ways and Means Chairman Bill Archer wants to eliminate the Internal Revenue Service, while colleagues like Dick Armey in the House and Connie Mack in the Senate want a flat-rate income tax, something the GOP’s Kemp Commission seems poised to endorse. Presidential hopefuls (including Bill Clinton) have unveiled or are developing their own tax plans. Flat taxer Steve Forbes, for example, has made radical change the centerpiece of his presidential campaign.

In revising the federal tax system, reformers should look at the evidence from the 50 states and the District of Columbia. Over the long term, there are very significant interstate variations in economic performance. For example, from 1960 to 1993 real personal income per capita rose more than 150% in South Carolina but only 60% in Delaware. At least part of the reason is that the 50 states differ widely in tax systems. New Hampshire has no sales tax and very little income taxation, while New York is near the top of the nation in its income tax burden, with substantial sales taxes as well.

Research on the relationship between state and local taxes and growth that I am doing for the Joint Economic Committee of Congress reveals several things. First, taxes matter a great deal. A state can dramatically improve its economic performance by lowering the overall tax burden. Tennessee has outperformed Kentucky over the long run because it has lower and better (from a growth perspective) taxes. The same is true of Washington state relative to Oregon; Texas or Florida compared with California; or Connecticut and New Jersey relative to New York. High taxes materially retard economic growth.

Second, the type of taxes matter. The statistical evidence generally shows that sales taxes do not seem to have much of a negative impact on growth. On the other hand, there is a striking negative relationship between income tax burdens and income growth. The experience of the states suggests that a move in the direction of taxing consumption as opposed to income would increase our nation’s wealth and prosperity.

Does this support a national sales tax as a replacement for the income tax? Not necessarily. Existing state sales taxes are low and do not cover many services. A national sales tax would require rates that are far higher (15% to 20%) and a base that is broader than any existing state tax, making it highly speculative to extrapolate from the evidence from the states.

That evidence also shows extremely strong cross-border activity on the part of price-sensitive consumers. Oregon, without a sales tax, has retail sales about 20% above what the national retail sales-disposable income ratio would predict, as shoppers from neighboring states take advantage of lower prices. With nearly half of the American population living in states adjacent to Canada or Mexico, this is a real concern with respect to a national sales tax.

What does the state and local experience say about a flat-rate income tax? The econometric evidence demonstrates that flat-rate income taxes lead to superior economic performance over variable-rate income taxes that raise the same amount of money (see chart). The 14 fiat rate states (some with zero rates) have markedly out-performed the 36 progressive rate states over the past 32 years. The same results are even more strikingly confirmed if one looks at population growth: On balance, people are migrating away from progressive tax states and toward flat rate ones.

Assume two states were otherwise identical in 1962. Both had income taxes raising the same revenue, but one was a flat rate tax of 4%, while the second state had a tax with a variable rate of 1% to 7%. A regression model suggests that personal income by 1994 in the flat rate state would be nearly 20% higher than in the progressive rate state.

Additionally, a national flat-rate income tax would reduce the incentives for the states to follow perverse fiscal policies that federal tax policies currently favor. The federal government rewards states that have high taxes by giving larger tax deductions to their citizens. What is worse; citizens can deduct relatively harmful state and local income taxes against their federal tax, but not sales taxes that are more benign in their economic impact. A true flat rate tax would end state tax deductibility, and in any case would reduce the existing bias against a growth-oriented state and local fiscal policy centered on low, consumption-based taxes.

On balance, the “states’ evidence” points to the need to lower the overall tax burden and to abandon progressivity in marginal rates. Economic growth arguments aside, polling data from the Advisory Commission of Intergovernmental Relations show that people consider the current federal income tax the most unfair of all taxes. My own rough estimate suggests that we now have an “army” involved in tax administration conservatively double the U.S. Army in size. On efficiency, equity and administrative grounds, the evidence supports radical change.

The current debate over minor tinkering in the tax code needs to refocus on more fundamental change. In doing so; looking at the diverse fiscal experiments under way in our federal system would help find a new tax system that promotes economic opportunity and justice at an affordable administrative price.