They never learn - educational test scores decline

National Review, Oct 28, 1988 by Ed Rubenstein

IN THE MID SIXTIES a group headed by Professor James Coleman, then of Johns Hopkins University, administered standardized intelligence and achievement tests to approximately 605,000 elementary- and secondary-school children in schools spread throughout all fifty states and the District of Columbia. Coleman and his collaborators had expected to be able to explain the great differences in average test scores between white and minority-group students by differences in the quality of their schools, as measured by such factors as average class size, number of textbooks, library facilities, age of school buildings, and teachers' education and background. He found, instead, that black schools did not spend significantly less than white schools and did not have larger classes, fewer or less adequate textbooks, less qualified teachers, etc., etc.

More importantly, the Coleman Report found no tendency for students from a given socio-economic background to perform significantly better in newer schools with high per-pupil expenditures than in older schools with low per-pupil expenditures.

These findings-although corroborated in subsequent studies done in this country and Europe, as well as in a second Coleman Report (1981), which found that, on average, students in private schools achieve more, even with large classes and fewer resources, than those in public schools-are still hotly disputed by the public-education establishment.

The most recent salvo was fired by Albert Shanker, who accused the Department of Education of fudging numbers in a chart that showed steady growth in elementary- and secondary-school expenditures on one side, and a steady decline in SAT scores on the other. Mr. Shanker complains that the expenditures were in current dollars, not adjusted for inflation, while the SAT scores, which go from 400 to 1,600, were displayed on a graph that went only ftom 800 to 1,000.

Unfortunately for Shanker's argument, the negative relationship between expenditures and test scores persists even after these changes are made, and is more pronounced when outlays are viewed on a per-student basis:

ELEMENTARY/SECONDARY PUBLIC EDUCATION

SPENDING AND ACHIEVEMENT

Per Average

Total (Billions) pupil, SAT

Current $ 1985 $ 1985 $ Score

1951 7.3 30.2 1,295 970

1963 21.3 75.1 1,971 980

1970 45.5 124.6 2,983 948

1975 70.6 138.4 3,329 906

1980 104.1 130.9 3,470 890

1985 147.6 147.6 4,051 906

After peaking at 980 in 1963, average SAT scores fell steadily to a low of 890 in 1980, after which they rebounded to 906 in 1985. Real per-pupil expenditures rose throughout this period, and in fact increased at a faster clip during the 1960s, when the decline in scores first became evident, than during the 1950s, when scores were generally on the rise. Interestingly, the rate of decrease in average test scores abated somewhat during the 1975-1980 period, just when the rate of increase in real per-pupil expenditures was also declining.

Public-school enrollment fell by more than 6.5 million students between 1970 and 1985, and the ratio of pupils to teachers fell from 22.3 to 17.9. We can thus say that the productivity of the average teacher, as measured by the number of students he teaches, declined by 20 per cent over this period. (This comparison is probably too charitable, since it ignores increases in non-teaching personnel, as well as the decline in SAT scores.)

The long-term decline in educational productivity contrasts sharply with increased output per worker in the manufacturing sector of the economy:

PRODUCTIVITY TRENDS IN

EDUCATION AND MANUFACTURING

Pupil/ Output Per

Teacher Worker

Ratio (1977 = 100)

1960 25.8 61.4

1970 22.3 80.0

1980 19.0 99.9

1985 17.9 124.9

The productivity of manufacturing workers has more than doubled since 1960, while that of teachers fell more than 30 per cent. Both teachers and manufacturing workers clearly have more productivity-enhancing equipment to work with now than in 1960, but the public schools, unlikemanufacturers, face no meaningful competition, and thus have no economic incentive to pass the benefits of higher productivity through to the taxpayers.

OF COURSE, teachers might argue that they are able to provide considerably more individualized attention in today's smaller classes than their counterparts in 1960, and that the quality of the learning experience has improved. This assertion is belied by the decline in SAT scores, as well as a recent Department of Education report showing that average class size in Japan is approximately 35, about twice the U.S. level, and yet Japanese students do much better than Americans in math.

In 1950, public schools educated 16.6 per cent of the U.S. population at a cost of 2.2 per cent of the nation's GNP; in 1985 it took 3.7 per cent of a much larger GNP to educate the same proportion of the population. The reasons behind the test-score decline are obviously more complex than just the financial resources being applied to education.