Great supply-side debate: do marginal tax cuts work?

The central tenet of supply-side economics is that Americans change their behavior to skirt taxes, like children who change their route home from school to avoid a bully.

Reduce the tax threat, supply-siders say, and taxpayers will take a direct course to more productive behavior -- substituting saving for spending, working for leisure, starting a business for playing the money markets.

The only problem is, nobody can prove it.

''It is important to realize we do not yet understand the potential long-term impact of a society with heavy marginal taxes, or of one with more moderate marginal taxes,'' says Stanford economist Michael Boskin.

And the central tenet of supply-side opponents is that the supply-siders won't be able to prove it because they're wrong.

''The best professional estimates are that some grains of truth are involved, '' says Lawrence Klein, a Nobel Prize-winning economist from the University of Pennsylvania. ''We can find some effects of tax rates on labor supply or on savings, but they are small, slow, and sometimes not statistically significant. This is hardly a basis for forming a grandiose scheme of public policy.''

Monetarists, for instance, have their faith well documented in the massive tome ''A Monetary History of the United States,'' produced by economists Milton Friedman and Anna Schwartz. Supply-siders are just beginning the number crunching needed to create such underpinnings for their theories.

''Boskin is right on,'' says James Gwartney, a Florida State University economics professor. ''What we need to do is perfect our ability to predict the effect of marginal tax rates.''

Dr. Gwartney, in a paper delivered at a supply-side economics conference sponsored by the Atlanta Federal Reserve, presented some preliminary figures on tax avoidance caused by high rates.

In 1965, Gwartney figures, taxes bit off 19.5 percent of a worker family's gross income (defined as 1.5 times the average manufacturing wage). By 1980, federal, state, and local increases, together with inflation-caused ''bracket creep,'' had shoved the tax take up to 35.13 percent.

This increase can ''induce people to substitute leisure for work, activities that don't produce revenue for those that do,'' he says.

Gwartney admits statistics don't show that rising taxes make the head of the household work less. But he says there's a significant effect on secondary workers. Working wives, for instance, decide the second income isn't worth it when they look at the bottom line of the joint 1040 tax return.

And he says high rates definitely create a bull market in tax shelters -- side businesses with artificially arranged losses that can be deducted from income.

Analyzing joint returns featuring sidelines sometimes used as tax shelters -- partnerships, small businesses, real estate -- Gwartney found 14.2 percent reported deductible losses in 1966. In 1978 the number jumped to 31.2 percent.

''This is very consistent with the idea (that) rising marginal tax rates rechannel work effort,'' Gwartney claims.

Critics, such as Dr. Klein, say the number simply reflects harder economic times.

Reversing this trend, claims Gwartney, requires a tax cut larger than the package passed last year.

''For most people, the legislation will not lower marginal tax rates,'' he says. The cuts will be offset by inflation and increases in state, local, and social security taxes for all working couples earning between $20,000 and $80, 000, he says.

Gwartney would have preferred ''a smaller tax cut that would have been indexed.''

Michael Boskin, at the same conference, discussed whether high taxes discourage Americans from saving.

''It is clear a reduction in marginal tax rates on the return to saving will increase private savings,'' Dr. Boskin says. ''Not as rapidly as official administration forecasts, but faster than pessimists who say there is no response at all.''

He figures that a 1 percent reduction in the tax on income from savings instruments will increase the money poured into the US personal savings pool 12 to 15 percent.