Call it a liberal counterattack. The argument: The welfare state has reduced poverty, improved economic security and provided health care for millions who would otherwise lack protection, all without reducing savings and investment or threatening the future fiscal soundness of the nation.

The counterattack comes in a new book, "America's Misunderstood Welfare State," in which the authors assert that the country's social welfare programs "are taking a bum rap" from conservative critics.

"It is not the case that our social welfare budget is unaffordable, either in the sense that we are unwilling to pay for it or because it is driving us to economic ruin," say the authors, Theodore R. Marmor, professor of public policy at Yale, Jerry L. Mashaw, professor of law at Yale, and Philip L. Harvey, an attorney in New York.

They say a major reason the notions they consider false came to flourish is that in the 1970s the proportion of Americans in poverty stopped falling just when new welfare programs like Medicaid and food stamps were beginning to expand.

That, the authors contend, provided fertile ground for a superficial argument that the "war on poverty" was doing little good and that programs might even have pernicious effects on the poor that cause them to shirk work.

The real reason for the stubborn persistence of poverty, the authors say, is that starting in 1973, the performance of the economy has been markedly worse than from 1945-1973, with smaller productivity increases, high unemployment, two deep recessions and "greater income inequality."

These developments were the real causes of the problems now attributed to the welfare state, the three authors argue, and the social support systems moderated the impact of these economic changes on poverty, but could not overcome it.

In their economic-social programs analysis, the authors point out there have been more and deeper recessions since 1973 and that there has been no period of special growth (as during the Vietnam War in the pre-1973 period).

"During quarters in which the economy was neither dragged down by recession nor artificially stimulated by wartime mobilization . . . growth rates averaged 4.2 percent both before and after 1973." If the cost of social programs were the main factor in the post-1973 slowdown, it would have been reflected in non-recession periods too.

They also point out that over the past 30 years, other industrial countries had far higher increases in expenditures for social programs but nevertheless had equal or higher economic growth than the United States. Japan's outlays for social programs, adjusted for inflation, grew 8.4 percent annually from 1975-1981 and its gross domestic product grew 4.7 percent annually. In the United States, social program outlays rose 3.2 percent annually and gross domestic product also grew only 3.2 percent.

The authors say research simply has not corroborated the notion advanced by economist Martin Feldstein and others that the Social Security system reduces savings needed for investment, because people figure they'll be taken care of in old age. In fact, they say, some research suggests that Social Security causes people to save more.

The authors also say the notion that social program expenditures are uncontrollable is disproved by the fact that when economic growth rates slumped after 1973, social program growth began to slow correspondingly, in part because of restraints deliberately imposed.

The principal reason for poverty, the authors argue, is unemployment: Poverty rises every time unemployment does. And they see unemployment as an economic problem, not tied to the so-called social deficits -- sloth, illegitimate births, lack of education -- cited by critics of welfare.

As for the argument that Social Security costs eventually will go through the roof, the authors note that under the best projections, the cost of Social Security and Medicare will go up from 6 percent of GNP today to about 10 percent in 2060; that increase is no greater than occurred from 1965 to 1986.

But why pay Social Security and Medicare benefits to people who are not poor? Why not means-test the programs and save money?

The authors reply that Social Security and Medicare are based on a better, more humane approach -- preventing poverty rather than dealing with it after it happens. A huge proportion of the elderly would be poor except for Social Security: More than half get at least half their income from Social Security; about half would be under the poverty line but for Social Security and similar government cash payments, instead of only 11.4 percent.

Moreover, they argue, means-testing would undermine political support and viability because the public is not generally supportive of welfare programs.

Rather than means-testing the system to save money, a better approach, they argue, would be to tax more of the Social Security benefits (only half is now taxed and only for higher-income people), adjust retirement age rules upward a bit and look toward keeping more elderly and disabled in the work force.