The correlational evidence often cited to justify spending most resources on an early intervention-for-prevention strategy is open to alternative interpretations. Numerous studies showing that poor readers in the early elementary grades are likely to eventually drop out (Howard and Andrew, 1978; Lloyd, 1978; Kelly, Veldman, and McGuire, 1964) have been used to justify the heavy allocation of available compensatory education funds in the early elementary grades. For example, Chapter 1, which receives over 6 billion federal dollars each year, is primarily an elementary-grades compensatory education program, since 90 percent of students receiving assistance are in grades K through 6 (Slavin, Karweit, and Madden, 1989). But these correlational data probably do not mean that if we successfully teach each child to read in the first 3 years of school that the problems of later failure will be solved. The correlation between school failure in the early and later grades can also be interpreted as due to a continuing internal or external variable, call it "poor school climate" or "poverty," that contributes to low school performance at each stage of education. If the same problems that cause reading difficulties in the early grades persist as barriers to a student's learning in later grades, an investment strategy that attacks only the initial problems will not be adequate to the task. There is no early "inoculation" that will prevent later problems if most students' learning problems are caused by persistent undernourishment from the school or out-of-school environment. This situation will require continuing "booster shots" of extra assistance to students and schools to establish and maintain a normal healthy learning development across the years.
Of course, successful school experience in the early grades can make major contributions to the next stage of learning. Initial success builds a student's self-confidence and motivation for the next challenge, and basic reading and numeric skills form the essential foundation for most later learning activities. We are not arguing that resources for these efforts be depleted in favor of programs directed toward older students. But there is no scientific evidence that it is too late after the early elementary grades to remediate reading or math deficiencies or to motivate students through stimulating learning experiences. It is certainly worthwhile to provide sufficient resources to build a foundation of basic skills and self-esteem in the early grades, but there is no scientific justification for giving up on students who do not achieve early academic success or for assuming that early success will be sustained for all students without continuing extra help to them and their schools in the later grades.
In fact, research studies of compensatory education have repeatedly found a "fade- out" effect -- the cognitive learning gains achieved in the early grades quickly dissipate and eventually disappear entirely when there is no special program to follow through in the later grades on the initial investments (Natriello, McDill and Pallas, 1990). It is hard to justify an early-intervention-only strategy if a few years later the students who received the early assistance are indistinguishable in academic skills from comparable students who did not participate in the initial intervention program.
Research does not now exist that evaluates the effects of different resource allocation strategies across the grades. Future studies might explore the costs and benefits of different strategies to determine the extent to which increasing investments in preschool interventions such as Head Start makes more sense than adding funding for specific programs in the early elementary grades or extending extra academic and social services for disadvantaged students through the middle and high school grades. Obviously, the impact of increased investments at any stage of human development depends upon the effectiveness of the particular programs to be supported, and different kinds of assistance may make sense at different ages because the primary sources of motivation change as students grow up. Acknowledging these points, we urge that compensatory education be continued beyond the early years so that extra help is sustained for students who need it at each stage of development.