Politicians who talk about the family or the needs of children usually do so in order to plug their own pet projects. That's what you probably would have heard a few years ago if the House's Select Committee on Children, Youth and Families had been set up then. Its chairman, George Miller, is a liberal Democrat from the San Francisco Bay area; the ranking Republican, Dan Marriott, is from conservative Utah. But the committee's first hearings--despite some predictable theatrics--suggest that these politicians may be less interested in old panaceas than in forging a new consensus.

One basic fact emerged: American children today do not have the same problems as children of 15 or 20 years ago--because they are not the same kind of children. Democratic and Republican witnesses both pointed out that today's children are poorer, relative to the rest of society, than the children of the 1960s. One-fifth of children live below the poverty line, and 21 percent in single-family households. Why? Because in the 1970s poor Americans tended to have more children than rich and middle-income Americans. In contrast, during the baby-boom years, from 1947 to 1962, middle- and upper-income Americans tended to have almost as many children as poor people.

That presents us with problems that are new to us--but not at all novel in American history. Through most of our history the poor have had more children than the rich; the average child, at any given time, lived in a lower income household than the average adult. The baby boom years, with their high birth rates generally and especially among the affluent, were an aberration, not the norm, for a developed nation.

So, in every generation but the most recent one, it has been the task of American families and schools to improve the lot of children. We have taken the below-average income children of one generation and made them into the above- average adults of the next. Each generation of children has, in effect, leapfrogged its parents-- and in the process created vast economic growth and widespread affluence.

American families and schools faced a somewhat different task in the 1960s and 1970s. In a time of great economic growth, an unprecedentedly high percentage of children began life's journey already at high levels of affluence. They seemed to need not economic improvement, but enrichment; the task was not to train people for production, but to train them in the arts of consumption. Adults, too, decided to take advantage of their affluence and their increasing freedom from traditional constraints. Among the results were lower birth rates and more divorces (and remarriages), more spending on leisure, and less money saved and invested. Also among the results were the educational practices that the recent National Commission on Excellence in Education, joined by a virtually unanimous national chorus, has decried.

Now, all these trends work against the goal of helping today's less affluent children improve their economic condition. Single- parent households, two-paycheck families, larger numbers of illegitimate births and births to teen-age mothers--all these make such improvement less likely and more difficult. Facing these problems forces politicians to abandon pet projects. The political left distrusts the moral agenda of the right, but liberals have to admit that changes in family patterns--especially the increase in single-parent families--hurt many children's chances in life. And even Miller is not at all sure that government day care is the answer.

The political right dislikes expensive government programs; but conservatives have to admit that if many of these children's chances are not somehow improved, the quality of life for everyone will deteriorate. How many conservatives feel sure President Reagan's solutions of school prayer and tuition tax credits are enough? From that uncertainty may come the makings of consensus.

No one is quite sure how to help children whose parents by definition have fewer resources --economic and otherwise--than average; yet it's in society's clear interest to do so. So perhaps the right may be willing to provide nutrition to young mothers and children, as the WIC program does. Perhaps the left will admit that it would be helpful if so many babies weren't raised by single teen-agers, if parents didn't divorce quite so often, if fathers did not abandon family responsibilities so often, if the social work profession did not do so much to discourage adoption.

Adoption, as Miller points out, is a way to help improve the economic status of many children; birthparents who give up children for adoption tend to be very young and relatively poor, while adoptive parents tend to be older and more affluent. There is plenty of demand, as the increase in numbers of adoptions of foreign children show, yet barriers keep thousands of available children from being adopted.

In education and in family policy generally we seem to be moving toward a consensus that the policies of the 1970s, designed to liberate an affluent body of parents and children, are inappropriate for the 1980s, when the task is to help a notably less affluent body of parents and children to improve themselves and the nation. Both left and right may be ready to agree that we no longer have the luxury of assuming that children will automatically leapfrog their parents' generation. Locally, politicians and ordinary citizens have been working toward solutions; now maybe national politicians can define the problems and articulate that consensus.