It has been a little more than two years since President Clinton signed legislation ending welfare as we knew it. What can we conclude so far beyond noting that substantial numbers of families continue to leave welfare and that legions of children (despite what critics had predicted) are not starving in the streets? Let me make two broad points that, on rapid reading, might seem inconsistent, though they are not.

The first has to do with the (now) obvious timidity and lack of faith with which great numbers of politicians, analysts, social workers, journalists and others previously viewed welfare and possibilities for its serious reform. Or, more pertinently, it’s striking to now note the degree to which such inside players had underestimated the ability of people to escape welfare successfully if only they were pushed and poked adequately to do so.

The second point has to do with noting the limits -- as well as the not always comfortable meaning and implications -- of such pushing.

A good place to enter the discussion is a speech last July by Mayor Rudolph Giuliani in which he vowed to end welfare in New York City, and to do so "completely," by the year 2000.

"From the welfare capital of America," he said, "we will become the work capital of America, the place that understands the value of work in a deep and metaphysical sense much more than any other place in the United States."

How can one not be impressed by such an audacious politician, one with such implicit confidence in the ability and conscientiousness (latent though they may be sometimes) of just about everyone? For that matter, how can one not be impressed with a public servant who uses words like "metaphysical"?

But how realistic is Giuliani’s promise to end welfare completely, keeping in mind that alcoholics and drug abusers, for instance, would be expected to work just like everyone else?

It's not realistic at all. My guess is Rudy had an overheated, out-of-body experience on the day he said what he said last summer, fantasizing that he wasn’t in Manhattan any longer, but 3,000 miles west in Disneyland. In this one instance, I essentially agree with his critics, including the head of an addiction treatment center in New York, who said that most people in such programs are "really extremely broken down and are not people who are either psychologically, emotionally or physically in any condition to work."

Still, the larger point to be made here is that we should welcome and celebrate Giuliani's imagination, and as well as that of other politicians (most certainly Gov. Tommy Thompson of Wisconsin) who have fundamentally changed expectations about the capacity of recipients, in fact, to leave welfare safely.

For methodological and other reasons, I don’t necessarily trust every statistical claim in this area, and one must also acknowledge the exquisitely-timed benefit of a strong national economy over the last half-dozen years. But did anyone really predict, back in 1993 when Giuliani was first elected, that New York City welfare rolls, within five years, would drop by more than a third? Or that much the same would hold true for the country as a whole? Or that welfare caseloads in Wisconsin, according to a scholar whom I do trust, would drop over the last decade by about three-quarters?

The lesson here is that we have needed to be bolder than we had been -- not colder, but bolder -- in reining in welfare and eligibility for it. We have needed to do so not just in the interest of public treasuries, but much more to the heart, on behalf of millions of families: mostly mothers and children who have been enwrapped, not in compassionate public arms, but in a well-intentioned dependency that not a single person reading this column would wish on anyone he or she loved.

What, though, is the "deeper" idea, as the mayor of New York City might put it, behind the kind of "pushing" we have been talking about? What is the basis for believing that poor people, if they are to be better served by both public and private agencies, need to be simultaneously helped and hassled?

Lawrence Mead, a New York University political scientist who consistently writes with insight and courage about poverty, calls that idea "paternalism," by which he means the right of society, at least in some respects, "to tell dependents how to live."

He compares this approach to that of "traditional" welfare policy, which defers "to the capacity of clients to live their own lives," to make almost all of their own decisions about going to work, getting an education, etc. Paternalism, Mead says, (or "tough loving," as others might put it more colloquially) assumes that those who receive services such as welfare "need direction by others."

He is right, of course, as a significant proportion of people who come to depend on public assistance, or homeless shelters, or drug treatment programs do so, blunt as it sounds, because they have misused their freedom.

At the very same time, it's hard not to be made uncomfortable by the fact that we are talking about treating adults like adolescents. Sure, it's good that welfare policies have come to be sterner in requiring grown-ups to fulfill grown-up obligations. There is dignity in that. Still, the means by which we increasingly seek to assure responsible behaviors by recipients make a part of me cringe, as I have a hard time considering them without recalling being yelled at in school. Mead puts this unease in less personal and visceral, but still acute terms when he writes:

"Paternalism, after all, means government oversight of the personal lives of dependent adults. In the land of the free, this is not welcome."

-- Mitchell B. Pearlstein is president of Center of the American Experiment, a conservative think tank in Minneapolis.