THE WASHINGTON budget debate, like such debates for a decade, has been styled as an argument between heartless accountants and compassionate doers of good.

Do we care about the poor? The handicapped? The unemployed, the unwashed, the helpless? When you put the question that way, it is easy enough to portray those seeking to hold down spending as the heartless accountants.

But a decade after the explosion of social programs known as the Great Society, isn't it time, finally, to ask a different question: Do the programs work? Has federal aid to education made a significant difference in the nation's schools and helped poor children catch up with their middle-class schoolmates? Have federal urban development programs improved the country's big cities? Has the Law Enforcement Assistance Administration made the streets safer? Has the Comprehensive Employment and Training Act given the hard-core unemployed the training they need to land jobs?

The Health, Education and Welfare Department's education budget, which now totals about $12.5 billion, is a case in point. The Elementary and Secondary Education Act (ESEA) was created in 1965 as a centerpiece of Lyndon Johnson's Great Society to pump out money for poor children. The assumption-right or wrong-was that extra "compensatory" instruction would help "disadvantaged" pupils learn at average rates and thereby break the cycle of poverty.

Although some scattered evidence is beginning to emerge suggesting there has been at least some impact on learning, for the most part it hasn't happened. The aid has grown steadily over the years, so that 80 percent of elementary schools end up with funds to help these students under the main program, Title I. About 6.5 million children are enrolled at least part-time in compensatory education at an average extra cost of $450 per child. But largescale, federally funded studies have generally been unable to find that the extra money has had much impact in improving education.

The major evaluations through 1973 showed that children in compensatory classes were achieving no more, and often less, than similarly disadvantaged pupils who remained in regular classes. In 1974, Congress asked the National Institute of Education to conduct a three-year, $15 million study of Title I. Its final report, submitted to Congress last year, simply sidtracked the question of whether the money had any impact in improving education. Instead it dealt with such other important questions as what types of compensatory programs worked best.

As one witness told a House education subcommittee last year, ESEA was an "unquestioned success." He was, literally, correct.

No follow through

The impact of federal education aid is more dubious in other offshoot programs. In 1968, Congress, again at Johnson's urging, established the "Follow Through" program. This was designed to continue the supposed learning gains made by poor children in "Head Start." The Head Start program served preschool children, while Follow Through was to serve those in the first three grades. But the tight budget brought on by the Vietnam War persuaded the Johnson administration to scale down Follow Through and make it instead an experiment in compensatory education.

University education departments and social science research firms got contracts to set up compensatory education projects in selected communities. Some emphasized drills in the basics, others tried to improve "self-concept" as the way to help students learn better.

The Office of Education also put out a contract to evaluate the efforts by compairing Follow Through students to similar poor students in: nearby schools. The evaluation, which lasted eight years and cost about $30 million, was released in 1977. It found that in eight of nine projects, students in Follow Through programs, which cost an extra $360 per child, gained no more or less than comparable students in regular schools. No type of project was found consistently successful.

"Apparently, it is genuinely hard to raise the test scores of disadvantaged children, even in comparison to those of other disadvantaged children, by means of the compensatory interventions that Follow Through has tried," said the evaluation report by Abt Associates of Cambridge, Mass. The budget for Follow Through has been $59 million a year. The Carter budget recommends another $59 million next year.

Also in 1968, Congress added a bilingual education program to ESEA. This was for "demonstration projects" to find better ways to teach English to children who come to school solely or chiefly speaking another language. By 1977, the bilingual education budget was up to $150 million, and it was still fraudulently called a demonstration program. Nothing was being demonstrated.

A controversial evaluation was done by the American Institutes for Research of Palo Alto, Calif. - and disputed by many bilingual education supporters. But whether or not Hispanic students in the federally funded bilingual classes were gaining a bit less in reading English than other Hispanic students, as the study said, the program certainly has not done what it set out to do: demonstrate the value of the bilingual technique. Still, the administration has recommended that the bilingual education budget go up another $15 million, to $174 million.

Said HEW Secretary Joseph Califano in announcing the new budget: "Our inept administration [of the bilingual education program] should not be used as an excuse for cutting it." Maybe not. But in what year - or decade - can we begin to judge the program on its merits?

The "innovations"

Washington has also given hundreds of millions of dollars to school districts to "reform" education by funding "innovative" projects. This effort rested on as least two assumptions - that "innovative" projects would be an improvement over existing offerings, and that they would also spread to other schools. Here, as elsewhere, the overriding assumption was that more money would make for better education.

Last summer, the Rand Corp, released its eighth and final volume of an evaluation of these projects, which covered reading, vocational eduacation, bilingual education and other efforts. It concluded that the assumptions were plain wrong. Other federal evaluations have "revealed inconsistent and generally disappointing results" as well. Rand notes."Although federal support for local school services has become well established, the 'decade of reform' that began with ESEA has not fulfilled its expectations."

Why? Educators, to the extent they discuss the problem at all, seem genuinely puzzled. The mystery of how to improve education from outside remains, after 14 years, still a mystery. But Rand offers one cogent suggestion as to why federal aid has had such a minimal impact:

"More money supplied by federal funds did not necessarily purchase those things that mattered; it did not buy, for example, more committed teachers, more effective project directors, more concerned principals and so on."

What did determine the success of a school, not surprisingly, was "not the amount of funds available but the quality and behavior of the local staff." Money for innovative school projects come mostly from ESEA's Title IV, whose budget was $197 million last year; the Carter administration recommended another $197 million for the coming year.

Guarding the status quo

All of this is by no means unique to education.The process is repeated over and over as federal social programs are first created, then expanded. A national social problem is identified, usually through the media, and Congress creates a grant program to do something about it. For law enforcement, it was crime in the streets. For job training, it was rising unemployment. For housing and urban development, it was decaying cities. And each year, at budget time, the question comes back to the seriousness of the problem - not to the quality of the solution.

Its not hard to figure out why. There is a trial of federal money - billions of dollars - stretching from here to Guam. It goes to states, counties, cities and thousands of institutions, from Ivy League colleges to veneral disease clinics. Not surprisingly, those whose jobs and institutions depend on these funds don't want the programs threatened. They form a powerful lobby for the status quo. To them evaluators and evaluations are as popular as bees at a picnic.

But Jimmy Carter came to Washington promising "compassion and competence." Presumably, his administration, unlike the Republican ones preceding it, genuinely believed in the social programs begun during the Great Society era. But mich of the electorate also believed money was flowing from Washington like an open fire hydrant on a summer day. The Carter administration would supposedly bring hard-nosed management to Washington, see that the money was spent wisely and prudently.

To read the administration's fiscal 1980 budget is to see, however, that Carter has yet to crack the system. The budget justifications, "zero-based" or not, defend the budget totals by the depth of the problems rather than the demonstrated value of the programs. The Carter budget does try to reduce subsidies to the middle class in favor of programs "targeted" at those most in need. But, generally, it doesn't ask the hard question about which programs are accomplishing their aims, and it fails to challenge those Great Society assumptions that have been proved doubtful or wrong.

The administration's decisions in: the budget and elsewhere seem based far more on traditional interest group politics than on tough analysis, more on fond hopes than on results. At midterm, the Carter administration still seems to believe it is better to err on the side of compassion than on the side of competence. CAPTION: Illustration, no caption, By Geoffrey Moss for The Washington Post