The administration has advanced an exploratory study by the Research Triangle Institute as confirmation that its cutbacks in welfare aid were "extremely successful." Whether you buy that characterization or not depends on what you mean by success.

The study found that several months after welfare aid to some working recipients had been cut off in late 1981 and early 1982, only 15 percent had returned to the case rolls in the same county in which they were initially enrolled. Although welfare turnover is normally much higher than commonly supposed, that was still a significantly lower return rate than the researchers found in a comparable sample for the year before.

The recipients who were forced off the rolls had, on average, smaller families, smaller welfare benefits and higher earnings than the typical welfare family. The simple arithmetic of welfare benefit calculation will produce that result. And many states also changed their welfare benefit computations so as to protect families with lower earnings from the harsher effects of the Reagan rules. So you would expect that these families would be relatively able to cope with their losses.

The researchers couldn't locate most of the recipients whose welfare benefits were cut off. Perhaps they moved elsewhere to search for jobs or to move in with relatives. Some may have returned to welfare in other jurisdictions. But most of those whom the researchers were able to find reported that they were still working and earning an average of about $700 a month. That's not a princely sum, and taxes, child care and other work expenses take part of those wages. Still, the net is considerably more than most of these women would receive if they quit their jobs and returned to welfare.

Those former recipients with relatively high earnings have, at least, made a sound economic choice. Having lost both welfare benefits and Medicaid, they are now poorer than before, but they are still substantially better off than if they had settled for the very low level of income that welfare provides, especially to small families. About half have some medical insurance through their jobs. Others are still eligible for food stamps, although food-stamp benefits for the working poor have also been scaled back. As long as they or their children don't get seriously sick, they have responded sensibly, if unhappily, to the incentives provided them.

Is this success? Presidential assistant Robert Carleson says yes, because welfare costs were cut--at least temporarily--and because the people forced off welfare were "removed from dependency." Of course, if your single criterion for welfare success is saving money, you might as well close all the welfare offices. None of the recipients would then be "dependent." Hungry, maybe, but not dependent.

On the other hand, you might think success meant reducing human suffering, giving millions of children a better chance to grow up to be self-sufficient and well-adjusted adults and assisting their mothers to increase their own earning potential as well. In this case, you might find the Reagan policies shortsighted.

As the Research Triangle researchers are quick to point out, we know very little about how the families in their sample are faring. But articles in The Wall Street Journal and elsewhere have shown that, while few former welfare mothers jettisoned their jobs in favor of welfare, most of them were having to struggle very hard to care for their homes and children and keep their jobs as well. Shouldn't society give at least some of them a helping hand?

Certainly some curtailment of benefits to high earners can be justified. Welfare experts have long pointed to the very generous rles for reimbursing work expenses that allowed some welfare mothers to remain on welfare even with earnings that were clearly sufficient to support their families. Even though relatively few recipients benefited from these rules, the last three administrations proposed limiting work expense reimbursement in order to be able to afford better benefits for poorer recipients. The Reagan administration has simply pushed this logic to an extreme conclusion and pocketed the savings. Now--except in states where other rules counter the effect of the Reagan changes --recipients with relatively low earnings experience little if any gain from their work effort.

In deciding whether this is sound policy or not, attitudes seem to matter more than facts. At a symposium on work and welfare at the American Enterprise Institute last week, Barbara Blum, president of Manpower Development Research Corporation and former commissioner of the New York State Department of Social Services, observed that there is plenty of evidence that welfare recipients want to work. Moreover, as studies done by her firm and others have shown, many welfare mothers, given necessary help, can earn enough to significantly improve their families' well-being.

But, as AEI researcher Jack Meyers pointed out, society--which is willing to make far larger and more speculative investments in tax incentives for business or exotic weapons development--has been unwilling to make a relatively small investment in providing work incentives and opportunities for welfare mothers despite clear evidence that it could reap future benefits for both their families and taxpayers. Why does the nation prefer instead to reap small short- term savings--at the cost of much hardship-- from the welfare population?

Perhaps, as suggested by other participants, it's because--all evidence to the contrary notwithstanding--welfare mothers are looked upon as outcasts who both deserve and prefer their hard and demeaning lot. If that's the case, it is not investment but punishment that is in order. Well, the poor are receiving their punishment and, after a brief respite, the welfare rolls are headed back up. A funny sort of success.