How To Tell If The Poverty War Works

The following is an excerpt from a statement presented by Daniel P. Moynihan, Director of the Joint Center for Urban Studies, to Senator Abraham Ribicoff's subcommittee on executive reorganization. Ribicoff called Moynihan's plan for an Office of Legislative Evaluation one of "the most significant ideas" to come before the committee and said he would introduce it in legislation this January.

I think we trifle with the intelligence of the American democracy when we assume that it will never accept bad news, must indeed be fed a constant diet of good news about past programs, accompanied by forecasts of catastrophe unless new and even greater ones are enacted.

I assume that the problem of objective evaluation of urban programs must become even greater now that the Federal government is moving beyond its original concern to improve the physical equipment of cities towards an effort to improve the human beings who live in them. No one need be told that people are harder to rehabilitate than buildings, though we begin to learn that the process is expensive and frustrating in buildings as well. More to the point, most of the high yield social interventions in the lives of urban populations have already taken place. (Birth control programs for the poor being perhaps the one great exception.) The other kinds of programs that are now being proposed are almost by definition marginal.

One comparison will suffice. The introduction of universal, compulsory education in the 19th century gradually transformed American society. This was the great change. Present efforts to add a few years of schooling at one end or another, to extend the curriculum and improve facilities, while useful and important, cannot possibly have the impact of that first innovation. To the extent that the new programs are directed to persons who have trouble coping with the education system as it is, they will be even more costly and marginal in their effects. It is one thing to be concerned about a high school drop-out. It is another and very foolish thing to pretend he is promising educational material.

In the Dark

If in spite of this situation social policy makers were required to operate in the dark, so that however much or little impact their programs might have neither they nor anyone else would have anything more than an intuitive, craftsmen's judgment of what worked and what did not, there would be nothing much else to say.

However, this is simply no longer the case. Over the past generation there have been enormous advances in the capacity of the social sciences to measure social change induced by public policy. These derive in great measure from the introduction of the high speed computer, but also from solid advances in our understanding of social processes themselves. "We live in a society," Otis Dudley Duncan writes, "that is understandable but not understood." I am sure Professor Duncan would wish to leave plenty of room for the eventual mysteries of life, but it is fair to state that the first order effects of many social programs are now subject to rudimentary, but nonetheless useful measurement....

Of all the major social initiatives of recent times, none, I think, arose more directly from academic research and evaluation than did the War on Poverty. Much of the basis for the program, for example, was laid by the careful economic analysis carried out in the 1950's by the Subcommittee on Low Income Families of the Joint Committee on the Economic Report. This attachment to research and evaluation was carried over into the operations of the Office of Economic Opportunity itself under the courageous and forceful leadership of Mr. Sargent Shriver. Surely one of the most attractive products of this process was Project Head Start. It was begun on the basis of good research information, and proceeded to research itself.

However, by the end of its second year of operation this research had pretty much demonstrated that while the program has strong initial effects on children during the summer, these effects are rather quickly attenuated as they arrive at their regular schools in the fall. This has led to suggestions that the program be year round, which is reasonable. But then, last month, Mr. Shriver announced yet a new proposal, Operation Keep Moving. The first point in this program is an across-the-board reduction of pupil-teacher ratios in elementary grades to that of Head Start, i.e., 15 to 1.

It will be clear that this is an expensive proposal. The current pupil-teacher ratio in public schools is 24.6, the current expenditure per pupil in average daily attendance is $532. To reduce the ratio to 15 to 1, in rough but representative terms, would require another $300 per pupil. This would represent an increase in expenditure of $1020 per year for a family with three children, or, as an example, 28.3 per cent of the median income of nonwhite families in the nation as of 1965.

No Effect

It happens that this is also a subject about which there is a body of some two hundred research findings, going back to the beginnings of empirical research in educational psychology and sociology almost half a century ago. In a paper presented to the American Statistical Association this past summer, Dr. Peter Rossi summarized these findings as follows: "By and large, class size has no effect on the learning of students, with the possible exception of classes in the language arts."

This past summer the Office of Education released its study, "Equality of Educational Opportunity" by Professor James Coleman, perhaps the second largest social science research project in history, which came to identical conclusions, stated as follows: "Some facilities measures, such as the pupil/teacher ratio...showed a consistent lack of relation to achievement among all groups under all conditions."

Now I would assume there is other research that contradicts these findings, and I know that the experience of individual teachers often does so. I would further assume that it is possible that a really sharp reduction in class size might make a very great difference. But it is also possible that a really sharp increase in teachers' salaries might have even greater effect. I don't know. The point is that in making proposals of this kind, which call for a massive allocation of public funds, we ought to get into the habit of reviewing the research at the outset and stating our case for moving contrariwise if that is our wish. How else is public confidence to be maintained?

The usual whispered argument, of course, is that to be candid about public policies that don't produce much progress is to give a weapon to the enemies of progress. This is an unworthy argument; there are never grounds for concealing truth about public matters. (As best the truth can ever be had.) But it is also an absurd argument. The American public supports a fantastic array of social services, and does so in ever larger amounts. The issue, then, is not whether, but which.

Thus with regard to persons living in or near to poverty, a fundamental issue about which Professor Rainwater spoke last week is to choose between a strategy of services, which Mr. Shriver's proposal would entail, as against a strategy of income. The amounts of money a Project Keep Moving would require are in the range of those that would be needed to establish a national family allowance. I can imagine a good argument being made that if there is an extra thousand dollars a year to go round for every family in the nation, or every poor family, that the best thing would be to give them the cash and let them spend it on things they think they need most, which might well be formal education for many, but is surely to be more varied than any formula laid down in Washington would permit.

In any event, to propose spending the money on services which such research as we have suggests will produce little or no effect is to risk being thought ridiculous or worse by members of the public, and we would delude ourselves if we did not see that this judgement has already been reached by large numbers.

Even more to be deplored than our tendency to ignore the results of research and evaluation in shaping public policy is our tendency to undertake great and promising enterprises of just this kind only to forget about them and allow them to wither on the vine as other more interesting subjects come along. Let me cite, for example, the research and evaluation programs that accompanied the demonstration projects funded by the President's Committee on Juvenile Delinquency and Youth Crime, established in 1961. As few programs in the history of American government, this one set out consciously and openly to test a set of hypotheses about the nature of deviant behavior in slum neighborhoods. It was a physiocrat's dream and attracted great attention in the press and in university circles.