At last, Democrats are allowing their sense of probability to inhibit their enjoyment of fantasy. Hitherto, many Democrats have said that Social Security has just two negligible difficulties: an immediate "cash-flow" problem solvable with judicious borrowing, and another difficulty due to arrive around the year 2010 but which may not arrive -- at least, assuming fortuitous mortality, fertility and productivity developments.

A fable: a politician and an economist fall into a deep pit with steep sides. The politician exclaims: "How will we escape?" The economist replies: "Easily. First, we'll assume a ladder." For frightened legislators, politics is the art of imaginative assuming.

But the balloon of fantasy has been punctured by two needle-sharp articles--the most important journalism of 1982--written for the New York Review of Books by Peter G. Peterson. In the 1960s, the Review tried to shock the bourgeoisie by printing on its cover a diagram for making Molotov cocktails. Now it has managed to publish something really hair-curling and--such is the irrelevance of the political left --the author is a Republican, the board chairman of the Wall Street investment firm of Lehman Brothers, and a former secretary of commerce.

Peterson refutes the hypothesis that demographics--the coming-of-age as wage earners of the postwar "baby boom"; the retirement of the relatively few Depression babies--will soon make the system solvent. That projection depends on assuming, among other things, "an unbelievable rate of sustained growth in productivity--about 3.1 to 3.3 percent per year from 1985 to 2005. That would far surpass any comparable period in U.S. history, even the boom years of the 1960s." Here are some productivity growth rates:

1948-67: 2.5 percent

1967-73: 1.6 percent

1973-81: 0.1 percent

Assuming productivity grows at even a rate of 1.9 percent--higher than in the 1970s (1.2 percent)--the annual deficit in the retirement and disability trust funds would exceed $100 billion by 2005, after which it would "explode," detonated by the retirement of the "baby boom." Add the hospital insurance trust fund, and the 2005 deficit exceeds $700 billion.

A 65-year-old retiring in 1982 who paid Social Security taxes throughout his career and was an average wage earner with a non-working spouse contributed a total of just $7,209 in payroll taxes during his working life. Given conservative assumptions about longevity and moderate assumptions about inflation, this retiree and/or his spouse will live to receive $520,000-- 75 times the dollar amount he contributed. If he began receiving benefits in January 1982, he received $803 a month, and got back the dollar amount of his lifetime contributions by September. Even adding interest (say, the Treasury rate, compounded) to the worker's payments, it would take him just three years and seven months to get back the amount he contributed.

Peterson notes that were it not for the growth of Social Security, federal revenues as a percent of Gross National Product would actually have declined between 1955 and 1980. Social Security spends more annually than all corporations invest in plant, equipment and research and development. Since 1949, average wages have increased 470 percent, average income taxes have increased 570 percent, average Social Security taxes have increased 6,480 percent. The maximum annual contribution made in the years 1937-49 was just $30.

This year, employers' share of payroll taxes will amount to a sum equal to about 50 percent of corporate profits. As a result, there is less investment, slower productivity growth and job creation --and an accelerating Social Security crisis generating pressure for accelerated Social Security tax increases that are part of the problem. Already, 25 percent of American workers pay more in Social Security taxes than in income taxes. Fifty-eight percent of the electorate is under age 45, and those voters are in the expensive family-forming, house-buying years. Sharply increased payroll taxes could depress the economy, deepen despair and ignite generational conflict.

Social Security's unfunded liability --the gap between the actuarially predicted costs of benefits and the taxes the beneficiaries are scheduled to pay--is $6 trillion. By the year 2030, America will be Florida writ large: the percentage of elderly in the population will be that in today's foremost retirement state. Of course, if Americans suddenly return to having five children, and life expectancy stops increasing, Social Security's crisis will moderate. But must biomedical advances against degenerative diseases be considered bad news because they undermine "optimistic" Social Security assumptions?

The Social Security crisis is evidence that the American government's heart is generally in the right place and that its head is frequently in the sand. But the sheer scale of the crisis should end, at least for now, the politics of imaginative assuming.