A study by the Congressional Budget Office indicates that President Carter's proposed welfare system would reach a far higher percentage of the population than has been thought.

The CBO estimates that if Carter's plan had been in effect in the high-unemployment, low income year of 1975, nearly a third of the 74.6 million households in the country would have been eligible for assistance.

But the CBO study also shows that, even under present law, a far higher percentage of the population is on the "welfare" rolls than is commonly understood.

The congressional experts say that about a fifth of the households in the country actually received some form of public assistance in 1975, and the number that took advantage of these assistance programs was not as great as the number eligible. Administration experts say Carter's plan might even reduce somewhat the population eligible for aid.

The CBO's total of households receiving welfare in 1975 was made up mainly of those receiving aid under three programs: Aid to Families with Dependent Children, or AFDC, which aids about 11 million persons, mainly needy mothers whose children are at home and husbands are not: the Supplemental Security Income (SSI) program, which aids the needy aged, blind and disabled and their dependents: and food stamps, which presently reaches about 17 million persons, some of whom are in the other two programs and some of whom are not.

These welfare programs that give aid to the needy make up only a small part of all government "transfer programs" under which money is given out to individual citizens.

In 1976, all government transfer payments came to $191.3 billion, and made up about one-seventh of all personal income in the country. The largest single transfer program was Social Security, which is now sending out nearly $90 billion a year to support, in whole or part, about 33 million people, or one-seventh of the population.

Other large programs include unemployment insurance and veterans benefits, and pension payments to retired federal employees.

Though "welfare" involves a larger share of the population than many people think, when measured alongside the total of all transfer payments it looks relatively small. Carter's whole welfare plan, embracing AFDC, SSI, the $6 billion a year now spent on food stamps, and assorted other smaller programs, would cost about $35 billion.

The CBO study says that, of 746 million "families" or household units in the country in 1975, 15.7 million [WORD ILLEGIBLE] about a fifth actually received assistance.

It says that 23.4 million or about [WORD ILLEGIBLE] per cent, would have been eligible for aid had the Carter plan been in effect.

About nine million of these would not have received direct welfare payment from the Treasury. They would have been aided instead through a special earned-income tax credit or welfare tax cut that Carter has proposed as a major part of his plan. This tax credit is one of several ways the Carter plan seeks to preserve an incentive to work, while giving expanded benefits to persons outside the work force.

Basically, Carter's plan involves merging the existing AFDC, SSI and food stamp programs into a single and perhaps simplified system of cash assistance to all needy households. The heads of some such households - a relatively small percentage - would be expected to work, and the government would provide jobs for those who could not find work on their own.

Benefits would decline as household income rises, another way the plan would seek to preserve work incentives.

A special House welfare subcommttee is now at work on the President's proposals, and hopes to have a bill written by the end of the year. But then it will have to go through three other House committees before it reaches the House floor, and the Senate awaits after that.

The CBO analysis, which was prepared for the House Budget Committee two months ago but was not widely publicized then, also gives some indication of how many winners and losers there would be under Carter's plan as compared to present law.

In 1975, it says, 17 million families, or 23 per cent, would have gained income, while 4.7 million, or 6.3 per cent, would have lost. The remaining 70.6 per cent would not have been affected (meaning their incomes would have changed less than $100).

The CBO figures indicate that:

Carter's plan would have cut the 8.3 million families below the poverty level in 1975 to 6.7 million, or 20 per cent. Forty-three per cent of those families would have gotten more money, and 27 per cent would have gotten less.

Carter's plan would have been more effective in reducing poverty among families headed by persons over 65. About 21.4 per cent of them would have gained income: 7.9 per cent would have lost.

It would have been more effective among single-parent families, particularly those with children under age 6. Some 52.1 per cent would have gained, 15.1 per cent would have lost.

The 1.2 million two-parent families with children and total income of less than $5,000 a year would have been reduced to 0.6 million.

Overall, the CBO analysis said, the majority of losers would have been families with incomes above the poverty line.

The greatest potential increase in numbers of families benefitted would have been among families earning $10,000 to $15,000 a year. Almost 1.3 million such families were helped, while almost 5.4 million would have been eligible under Carter's plan.

Most of the potential increase would have been due to Carter's more generous earned-income tax credit.

The tax credit also figured heavily in potential increases of 1.4 million families earning more than $15,000.

Those middle and upper income families, however, would not have been eligible for very large benefits. CBO figures showed that of the $14 billion more which the Carter plan would have paid out in 1975, $10.3 billion would have gone to families earning less than $10,000.

CBO warned that different economic conditions might make the comparisons for any other year look quite different. It also called its analysis tentative, because it used "imperfect data bases and computer models that are still in the developmental stage."