THE GOLDEN EGG theory of taxation, embraced by the economists of the supply side, holds that above all else it's the treatment of the wealthy that counts. These are the people, according to the theory, whose large and daring investments produce the golden eggs that enrich the whole society. But when these crucial few are oppressed by excessively heavy taxation, productivity stagnates and economic performance deteriorates. That, further according to the view from the supply side, is what has happened in the United States over the past decade.

Since President Reagan's tax reduction plan is based on this theory, it's no longer just another interesting hypothesis. The Reagan administration has not been altogether candid in its explanation of this plan, insisting repeatedly that it treats all income levels equally.

To follow the implications of the Reagan tax plan, it is essential to remember that the cuts would take effect over four years of continuing inflation. The administration assumes that prices will rise 35 percent over those four years. If a family of four had an income of $20,000 in 1980 and typical deductions, it would be in the 21 percent income tax bracket. If inflation follows the administration's expectation and this family's income stays exactly even with prices, its income in 1984 will be $27,000 -- the same real income, putting it, once again, in the 21 percent bracket. While the Reagan plan was cutting taxes for each bracket, inflation would have pushed this family up two brackets.

The same would be true for a similar family with $35,000 last year, paying taxes in the 32 percent bracket. By 1984, staying even with the assumed rate of inflation, it would have income of $47,250, which, with the same deductions and exemptions, would leave it right back in the 32 percent bracket. In both cases, because of the changing rate structure, these families would be paying a slightly lower proportion of their total income in federal income taxes. But in both cases most of that reduction would have been recaptured by the increases in the Social Security payroll taxes.

For people in the middle-income ranges, the promise of the Reagan plan is no more than protection against being pushed into much higher rates --as inflation would automatically do in the absence of a tax cut.But for incomes over $50,000 a year -- that is, for the top 5 percent of the income ladder -- the impact of the Reagan plan is very different. There the compounded effect of inflation and the new rates would produce substantial real tax cuts. A family with $80,000 last year and typical deductions would be in the 54 percent bracket. By 1984, with $108,000 -- the same real income -- under the Reagan plan it would be in the 43 percent bracket. At $200,000 a year, with the same assumptions, a taxpayer would drop from the 64 percent tax bracket to 49 percent. At that level the real cut from the taxpayer's 1980 tax would be worth about $17,000.

Your bracket is, of course, that rate at which you are taxed on the last dollar that you made. That rate is central to the supply-side theory, which operates in terms of the incentive to work, invest and produce. That incentive, the theory holds, is determined by the tax on the last dollar that you made -- or chose not to make. It's not an argument that has much relevance to people at average income levels. Few are in a position to save or invest much, and nobody responds sharply to a reduction of the marginal tax rate from 24 percent to 23 percent. But it's not unreasonable to suppose that much larger rate changes, for people with much more discretionary income, might make a significant difference in economic behavior.

It's curious, incidentally, that the advocates of the supply side should now be fiercely attacking Mr. Reagan for his failure to recommend immediate reduction of the top tax rate from 70 percent to 50 percent. The one thing that the Reagan plan will do absolutely reliably over these four years, regardless of any future inflation, is to bring that top rate down to 50 percent.

Is the Golden Egg theory likely to prove reliable? Is it worth trying? It's worth serious discussion, but the historical evidence is at best unconvincing. So far the argument for the Golden Egg rests more heavily on faith than on reason. It's difficult to show that cutting tax rates at the top affects the economy very differently from cutting them anywhere else.

Conventional Washington wisdom holds that Congress always moves faster to cut taxes than spending. But, as President Reagan has already perceived, his tax bill is running into much more serious challenge than his budget revisions. He is not, after all, proposing only to cut taxes. He is talking about the most profound change in the structure and distribution of American taxation since the end of World War II, and neither he nor his Treasury Department has yet begun to make a coherent case for it.