THE REAGAN administration's proposal to transfer responsibility for food-stamp payments T to the states is a serious mistake. With the cutbacks in other federal feeding programs, it could well lay the groundwork for a return to the disastrous situation of the 1960s, when 20 to 30 million Americans did not have enough money to buy a minimally adequate diet for themselves and their families.

In 1967, a medical team sent by the Field Foundation found areas of hunger and malnutrition in every part of the nation, rural and urban alike, and in every ethnic group. The worst cases were those who are always the most vulnerable: infants and young children, pregnant and nursing women, the elderly. Following that and similar findings of a 10-state nutrition survey, the White House Conference on Food, Nutrition and Health in 1969 recommended an expansion of America's large-scale nutrition programs.

Over the next three years, these programs raised the number of food-stamp recipients from 500,000 to 12 million and eventually to over 16 million. In 1977, another Field Foundation team, which included many of the same physicians who participated in the original study, retraced their steps. While in many areas (such as Appalachia) the facts of poverty had not changed, they found no medically demonstrable cases of malnutrition. In their findings, the doctors ascribed this elimination of malnutrition to the federal food programs.

Food stamps, in particular, have proven to be our prime bulwark against the hunger due to poverty. It is the program needy people depend on for three meals a day. Now, President Reagan would turn the program over to the states.

Despite all the insults heaped on it in the last few years, the federal government is much better equipped to deal with such massive, complex programs as food stamps than are most state governments. Very few have the strict financial procedures, the computerized record-keeping and the trained civil service needed to administer the program effectively.

By contrast, there is general agreement that the Department of Agriculture has been doing a very good job of managing the program equitably and efficiently. Although there have been charges of laxity and abuse -- and some can be substantiated -- the fact is that less than 5 percent of total spending goes to ineligible recipients or is over-issued. The rate of fraud is much lower than with income-tax returns. Incidentally, the major frauds generally have been perpetrated not by poor recipients, but by crooked caseworkers in collusion with disreputable retailers and hard-core criminals. This record of abuse could be improved, but not by fragmenting the program.

Overall, the states are the weakest link in our system of government. There is far more corruption, waste and favoritism at the "grass-roots" level than in the federal establishment. While the fact that something did not work in the past does not necessarily mean that it will not work in the future, we do not want to find ourselves once again in the situation where, as in the 1960s, eligibility for assistance might depend upon whom you know and even, in some cases, whether or not you are willing to wash the sheriff's floor.

It is by no means certain that the poor would receive equal protection under the law in every state. In the last 10 years, only two states increased welfare benefits enough to keep up with inflation.

The transfer of funding and administration raises the specter of bewildering, discouraging and expensive labyrinths of federal and local regulations for the different programs. It could well lead to a disparity in benefit levels like the ones that spawned the migrations from the South and from Puerto Rico after World War II and created the urban ghettos -- and riots -- in New York, Washington and Los Angeles in the 1960s. The most cogent reason to keep food stamps in the federal government is the nature of the program and the inability of the individual states, no matter how competent or willing, to assume the burden. Food stamps are basically an income supplement -- targeted specifically to food -- in an effort to ensure every American the ability to purchase an adequate diet during periods of economic distress.

The program has functioned well. For example, during the recession of 1973-74 the number of people on food stamps rose from 16 to 19 million, then fell back to 17 million with the recovery. As the present recession has deepened, the number has risen once again, this time to over 21 million.

Thus, in contrast to the costs of programs like Medicaid, which will rise predictably with escalations in the costs of health care, the cost of the food stamp program fluctuates with the health of the national economy.

Unlike the federal government, the states cannot maintain an operating deficit. Throwing the food-stamp program and the cost of its administration back to the states will mean that Michigan, for example, will face the highest cost at the very time when it is least able to pay for it. Unemployment there is running at 14.4 percent, well above the national average; tax revenues have dropped. Because there is relatively little federal spending in the Great Lakes region, Michigan gets back only 76 cents on every dollar sent to Washington.

Thus far, the food stamp program has been protected from such disparities. It is a national answer to a national need. It works. The middle of a recession that features high unemployment is not time to "fix" it.