President Reagan's "workfare" program for welfare clients was a failure when Reagan as governor tried it out in California from 1971 to 1974, according to a report issued by the state Employment Development Department a year after the experiment ended.

The report says that during the three-year experiment, only 9,627 persons were given community job assignemtns under the program, although California had nearly 2 million welfare clients at the time. The report also says that there was no evidence the program had any success in discouraging people from applying for welfare or in getting those already on the rolls to go off and find normal employment.

The report has relevance today because Reagan recently stated that hew wants to force all the states to adopt similar workfare proposals. Robert Carleson, who was Reagan's welfare director in California and drafted the current workfare proposals as a White House special assistant, disputes the conclusions of the report, contending that a substantial drop in the state's overall welfare rolls from 1971 to 1974 was at least partly a result of the workfare experiment.

Under the Reagan administration's workfare plan, welfare mothers and fathers would be required to work off part of their monthly welfare benefits in community work projects.

The California Work Experience Program, passed when Reagan was governor, required welfare clients to work 20 hours a week in public service jobs, such as directing traffic at school crossings or cleaning up litter. Another 20 hours was to be spent job-hunting. The individual didn't receive any pay for the work, beyond their normal welfare work-related expenses.

Reagan and his aides have often claimed great success for the program. However, the California report, published in 1976, gave these conclusions:

"CWEP as designated and implemented in California did not prove to be administratively feasible and practical." Only 35 counties participated in the program; 23 others didn't. Many counties disliked paying for child care, supervisory help and work expenses under the program. In 1974, the peak year of the experiment, only 182,735 persons were designated as potentially available, and of these, only 5,712 actually was assigned to workfare jobs. Only 886 lost benefits for refusing to work.

The program failed to demonstrate that the workfare concept would prevent people from applying for welfare by warning they'd have to work even if they got on the rolls. Welfare applications in CWEP counties actually showed a "significantly greater increase" than elsewhere.

The program failed to demonstrate that workfare pushed people who were already on the rolls to go off and get jobs. Actually, more people in non-CWEP counties left welfare to take jobs than in CWEP counties (the gap narrowed in 1974), but differences weren't statistically significant.

Carleson, in an interview, said the report made clear that many counties had entered the program half-heartedly and done a poor job administering it. Moreover, he said, "The goal was not to have people working in CWEP, it was to encourage people to seek their own jobs" by making it clear that whether they were off or on the rolls, "they would have to work."

But this measure, he contended, the program was a success because over the period 1971 to 1974 the welfare rolls statewide dropped about 300,000. Carleson conceded that much of that decrease was a result of other welfare changes. But he said he believed a significant part -- just how much he couldn't say -- was due to the sentinel effect of CWEP.

Similarly, he said the number of people leaving welfare for paid work statewide (not just in CWEP areas) had climbed from 15,667 in 1971 to 75,705 in 1974, and by his calculations -- which differed from those in the report -- the rate was 25 percent higher in CWEP counties. Again, he said it was hard to measure exactly how much CWEP contributed to the statewide numbers, but he believed the contribution was significant.

Carleson also challenged a recent study by the University of Chicago's Center for the Study of Welfare Policy. The study said welfare benefit and eligibility changes proposed by Reagan several weeks ago would cut any welfare and food stamp payments received by the working poor so disproportionately (when compared with a nonworking welfare client) that it wouldn't pay to work. Carleson said that this wouldn't be so if workfare was adopted, because then welfare recipients would have to work in any case and might as well keep their private jobs and make a bit more total money than straight-out welfare clients.

Carleson's view of this conflicts with that of Office Management and Budget press officer Ed Dale, who was quoted in published reports yesterday as saying that some of the welfare changes would have a work disincentive effect.