Chapter 7
Signaling, Incentives, and School Organization in France, the Netherlands, Britain, and the United States
JOHN H. BISHOP
Cornell University
Despite similar standards of living, the secondary education systems of France, the Netherlands, Britain, and the United States produce very different levels and patterns of achievement. In primary school, Americans do not trail their European counterparts. Reading ability varies little across these four countries. When 14 year olds were compared at the beginning of the 1980s, however, the French and Dutch were about 1.3 to 1.5 grade-level equivalents ahead of the Americans in math and science. At the end of secondary school, performance differentials were even larger.
What causes differences in secondary school achievement across these four nations? The first section of this chapter describes these achievement differences. Seven hypothesized proximate causes are evaluated in the second section. Four hypotheses can be rejected. The rest cannot: teacher quality, priority given to academics, student engagement, and time on task.
The third section addresses a more fundamental question: Why do American students, teachers, parents, and school administrators place a lower priority on academic achievement than their counterparts abroad? Why, for example, is student engagement in learning higher in France and the Netherlands? Some people blame American culture, antiintellectualism, or historical tradition. Such
ad hoc explanations cannot be ruled out, or in, by the analysis that follows. The purpose here, instead, is to propose an alternative explanation derived from economic theory and a few observations regarding the contrasting ways in which learning achievement is measured and then signaled to parents, school administrators, colleges, and employers in the five countries. The third section also shows how signaling theory, game theory, and agency theory provide a robust explanation of the learning deficits in American upper secondary schools.
According to the economic theory developed below, the fundamental cause is the structure of incentives for learning and high-quality teaching of demanding material. American employers reward credentials, but they fail to recognize and reward what is actually learned in school. Admission to the best colleges depends on measures of relative performance—rank in class and grades—and aptitude tests that do not assess what is taught in school, not external assessments of competence in particular subjects. Only one of the 50 states has a system of subject-specific external exams similar to the Baccalaureat (Bac), the General Certificate of Secondary Education (GCSE), and the Dutch exams. The result has been grade inflation and students selecting undemanding courses where it is easy to get a high grade. Students pressure each other not to study, in part because they are being graded on a curve. Teachers are pressured to keep failure rates low, so passing standards are effectively forced down by peer pressure against studying. The final section of this chapter summarizes the analysis and comments on the implications for economic analysis of education policy.
Differentials in Academic Achievement
The differences in achievement levels at ages 13, 14, and 15 are summarized in Table 7.1. The table presents data from studies conducted in the 1980s and 1990s comparing France, the Netherlands, England, Scotland, and the United States. The International Association for the Evaluation of Educational Achievements (IAEEA) studies sampled students at particular grade levels, not at particular ages. Consequently, age-adjusted scores on its tests are reported where possible and information on the age of the sample is provided in the footnotes of the table.
Reading. In the 1990–1991 IAEEA study of reading, age-adjusted scores indicated that American 9 year olds (see column 1 of Table 7.1) were reading about 58 percent of a U.S. standard deviation (SD) better than Dutch 9 year olds and about .20 SDs better than French 9 year olds. However, by age 14, differences between the countries (column 2) were tiny.
Mathematics. In the 1981–1982 study of mathematics achievement of 13 to 14 year olds conducted by the IAEEA, Dutch and French 13 to 14 year olds ranked second and third, respectively, behind only Japan. Of the 17 industrialized nations participating in the study of 13 to 14 year olds, Americans were
TABLE 7.1 Achievement in Lower Secondary School
|
1991 IEA Reading Age Adjusted |
1982 IEA Math |
1983 IEA Science |
1991 IAEP Mathematics |
1991 IAEP Science |
||||||||
|
|
|
|
|
Ages 14–15 (not adjusted for age) |
Adjusted for Age Difference |
Level Age 13 % Correct |
|
Level Age 13 % Correct |
|
|||
|
Age 9 |
Age Mean |
14 (SD) |
Ages 13–14 % Correct |
Mean |
(SD) |
|
Mean |
(SD) |
Gain Age 9 to Age 13 |
Mean |
(SD) |
Gain Age 9 to Age 13 |
France |
526 |
533 |
(68) |
53.9 |
- |
- |
- |
64.2 |
(20.3) |
- |
68.6 |
(17.1) |
- |
Netherlands |
494 |
523 |
(76) |
57.1 |
63.7 |
(16.1) |
62.2 |
- |
- |
- |
- |
- |
- |
England |
- |
- |
- |
47.1 |
55.9 |
(15.7) |
62.2 |
60.6 |
(21.4) |
29.8 |
68.7 |
(17.5) |
18.7 |
Scotland |
- |
- |
- |
48.4 |
- |
- |
- |
60.6 |
(20.3) |
26.5 |
67.9 |
(16.5) |
20.8 |
United States |
543 |
528 |
(85) |
46.4 |
53.7 |
(16.7) |
53.7 |
55.3 |
(20.9) |
25.4 |
67.0 |
(16.4) |
17.2 |
Columns 1, 2, and 3 are the age-adjusted means and standard deviations of the overall reading score in the IAEEA reading study (Elley 1992). Column 4 is a weighted mean percent correct for students in the grade where the majority have attained 13:00 to 13:11 years by the middle of the school year from the Second International Mathematics Study (McKnight et al., 1987). The French, English, and American students all had the same mean age, 14.1. Mean age was 14.0 for Scotland and 14.4 for the Netherlands. Adjusting for the greater age of the Dutch students would have lowered their percent correct by about 2 points. Columns 5 and 6 are the percent correct and standard deviation for ninth graders on the full 50-item IAEEA science test (Postlethwaite and Wiley, 1992). An estimate of how U.S. students would have performed on the full test was made by subtracting 1.1 percentage points (the average difference between core and full test scores for England and the Netherlands) from the U.S. core test score. The mean age of students differed a great deal. Mean age was 14:2 for England, 15.3 for the United States, and 15:6 for the Netherlands. Column 7 is an estimate of scores for the full 50-item IAEEA science test for students who are 15.3 years old, the mean age of U.S. students. The age gradient used was the average for Sweden (4.3) and Italy (7.4), the two countries for which it was available. Columns 8, 9, 11, and 12 are the mean percent correct and standard deviation from the 1991 IAEP study of mathematics and science achievement of 13 year olds (IAEP, 1992a, b). Columns 10 and 13 are the increase in the percent correct on items common to the tests given to 9 and 13 year olds. |
ranked twelfth, English eleventh, and Scots tenth (McKnight et al., 1987). After adjusting for small differences in mean age, American 14 year olds scored 10.7 points below Dutch students, and 7.5 points below French students, of comparable age (see column 4). The 1991 International Assessment of Educational Progress (IAEP) mathematics study obtained similar results (columns 8–10). The gap between French and American 13 year olds was 42.6 percent of a U.S. standard deviation (about 1.3 U.S. grade-level equivalents).2 British students were about halfway between the French and the Americans (IAEP, 1992a). The gap remained roughly constant even though the math achievement of 13-year-old Americans improved by .20 SDs between 1982 and 1992 (NCES, 1994).
The performance gap between the American and European students grows even larger during upper-secondary school (see Table 7.2). The Americans who participated in the Second International Math Study were high school seniors in college preparatory math courses, such as trigonometry, precalculus, and calculus. This very select group, representing 13 percent of American 17 to 18 year olds, got 39.8 percent of the questions correct. The 6 percent of English students studying mathematics at A-level got 59.8 percent correct (McKnight et al., 1987). Substantial proportions of French and Dutch secondary students specialize in mathematics and science; 20 percent of French youth are in the mathematics and science lines known as C, D, or E of the lycee general. The questions asked on their final examinations suggest that these students achieve at a very high level.
Science. In the 1983 IAEEA study of science achievement of 14 to 15 year olds, the Netherlands ranked third and the United States ranked last among 17 industrialized countries. After a rough adjustment for age differences, American students lagged slightly more than half a standard deviation (about 1.4 U.S. grade-level equivalents) behind English and Dutch students (see column 5, Table 7.2).
The 1991 IAEP science study found that at age 9 American students were ahead of students in Scotland, England, and most other European countries. Data for France and the Netherlands are not available for this age. By age 13, English, Scotch, and French students were ahead, although the differences were small and not statistically significant (IAEP, 1992b). The gap is smaller in the more recent study in part because overall science achievement of 13 year old Americans rose by .21 SDs between 1982 and 1992 (NCES, 1994).
TABLE 7.2 Achievement at the End of Upper Secondary School
|
1982 IEA Mathematics Final Year of Secondary School |
1983 IEA Science—Final Year of Upper Secondary School |
|
|
|
||||||||
|
|
|
|
Physics |
|
|
Chemistry |
|
|
Biology |
|
|
|
|
% Correct |
% Age Group |
% Time Math |
% Correct |
% Age Group |
Hr. Per Week |
% Correct |
% Age Group |
Hr. Per Week |
% Correct |
% Age Group |
Hr. Per Week |
Total Science Homework |
France |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
Netherlands |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
Belgium |
50.0 |
10 |
20 |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
Finland |
60.6 |
15 |
14 |
37.9 |
14 |
2.0 |
35.9 |
16 |
1.0 |
50.2 |
41 |
2.0 |
3.1 |
Norway |
- |
- |
- |
54.1 |
10 |
5.0 |
44.3 |
6 |
5.0 |
55.4 |
4 |
5.0 |
- |
England |
59.8 |
6 |
21 |
62.4 |
6 |
5.1 |
69.3 |
5 |
5.2 |
62.4 |
4 |
5.2 |
7.2 |
Scotland |
42.8 |
18 |
17 |
- |
- |
- |
- |
- |
- |
- |
- |
- |
- |
United States |
39.8 |
12 |
14 |
45.3 |
1 |
5.0 |
37.7 |
2 |
5.0 |
38.1 |
12 |
5.0 |
2.8 |
Column 1 is a weighted mean percent correct for students in the final year of secondary school from the Second International Mathematics Study (McKnight et al., 1987). The mean age was 17:8 for the U.S., 18:1 for England, 16:9 for Scotland, 18:6 for Finland, 19:2 for Sweden, and 18:3 for Belgium. Column 2 is the share of the age cohort in advanced mathematics courses included in the study. Column 3 is the share of school time spent in mathematics classes. Columns 4, 7, and 10 give the percent correct for students studying each science subject in the final year of secondary school. Columns 5, 8, and 11 are the proportions of the age cohort taking each science subject in the final year of secondary school (for the U.S. it is the share of students taking their second year of the subject). Columns 6, 9, and 12 are the number of hours per week spent in classes in each science subject (Postlethwaite and Wiley, 1992). The mean age was 17:5 to 17:10 for the U.S., 18:0 for England, 18:7 for Finland, and 18:11 for Norway. |
Few American upper secondary school students study science in depth (see Table 7.2). Only 1 or 2 percent of this age cohort takes two years of physics or chemistry. Despite the highly selected nature of this group, many of whom were taking the subject for advanced placement college credit, only 47.5 percent of the questions were answered correctly on the IAEEA physics exam and only 37.7 percent were correct on the IAEEA chemistry exam. The 4 or 5 percent of this age cohort of English youth who in their thirteenth year of schooling were studying these subjects for their A-level exams got 62.4 and 69.3 percent correct, respectively (Postlethwaite and Wiley, 1992).
Teacher Quality, Time, and Engagement: The Proximate Causes of Achievement Differentials
American elementary school students do not lag their counterparts in Europe. Indeed, in reading they are substantially ahead and in science slightly ahead (see rows 1 and 13 of Table 7.1). What, then, caused the large deficits in achievement in mathematics and science at the end of secondary school? Why does achievement lag in math and science but not in reading? Let us start by looking at seven proposed proximate causes of achievement differentials across countries:
- Diversity
- Restricted access to secondary education
- Teacher quality and salaries
- Overall spending per pupil
- Priority given to academic achievement
- Time devoted to instruction and study
- Engagement or effort per unit of scheduled time
The purpose here is not to select a single most important explanation for U.S. students lagging their French and Dutch counterparts. Rather, the objective is the more modest one of narrowing the list of possible causes.
Diversity
Non-Hispanic whites score about .45 GLEs higher than the overall U.S. average on NAEP reading tests, about .56 GLEs higher on NAEP mathematics tests, and .98 GLEs higher on NAEP science tests. If all French and Dutch students are compared to the 77 percent of American students who are neither black nor Hispanic, the European advantage is smaller. For mathematics at age 13, the gap would be about 0.9 GLEs in both 1982 and 1991. In 1983 white U.S. 13 year olds were about 0.5 GLEs behind the Dutch in science and in 1991 about .6 GLEs ahead of French 13 year olds.
But is it really fair to compare the non-Hispanic white population of the
United States to the total population of France and the Netherlands? The United States is not the only country with a diverse student population. The Netherlands accepted 120,000 immigrants in 1990—twice the rate of immigration into the United States. In both France and the United States the share of students who are taught in a language different from their mother tongue is 6 percent; it is 5 percent in Scotland, 12 percent in Canada, 15 percent in Northern Italy, and 20 percent in Switzerland (IAEP, 1992a). If scores are adjusted for the demographic and socioeconomic backgrounds of students, why not hold parent's education constant as well? If this were done, the French/Dutch lead over the United States would increase.
Access—Numbers of Students and Graduates
It is sometimes said that low achievement is the price that must be paid for greater access. However, only the United Kingdom exhibits the expected trade-off between achievement levels and enrollment ratios (see Table 7.3). Only 43
TABLE 7.3 1991 Enrollment and Completion Rates
|
France |
Netherlands |
United Kingdom |
United States |
Percent enrolled full time in secondary school |
|
|
|
|
Age 16 |
92.0 |
97.2 |
62.4 |
90.2 |
Age 17 |
86.4 |
90.0 |
43.1 |
74.7 |
Age 18 |
57.2 |
67.4 |
12.3 |
21.1 |
Age 19 |
31.6 |
41.5 |
3.4 |
5.0 |
FTE enrollment in tertiary education |
|
|
|
|
Age 18 |
19.1 |
12.7 |
24.4 |
33.1 |
Ages 18–21 |
26.6 |
19.5 |
16.0 |
33.4 |
Ages 22–25 |
12.7 |
14.0 |
4.8 |
13.5 |
Ages 26–29 |
4.0 |
4.0 |
2.2 |
6.2 |
FTE years in school between ages 16 and 29a |
4.6 |
4.9 |
2.3 |
4.1 |
School enrollment rate, Ages 5–29 |
57.7 |
55.2 |
52.7 |
55.2 |
Secondary diplomas awarded / population of theoretical completion ageb |
75.8 |
82.2 |
74.4 |
75.5 |
First-degree graduates from universities / population of theoretical completion age |
16.3 |
8.3 |
18.4 |
29.6 |
SOURCES: OECD (1993), NCES (1992), and Government Statistical Office (1992). a Calculated by summing the ratios of FTE enrollment to population for one-year age groups from ages 16 to 29. b The U.S. data do not include GED certificates. The labor market does not view the GED as equivalent to a high school diploma. GED-certified high school equivalents are paid 6 percent more than high school dropouts but 8 to 11 percent less than high school graduates. The graduation rate for the United Kingdom is spuriously high because it counts regular GCSE exams taken at the end of the eleventh year of schooling as graduation. If one or more A-level exams had been the definition of secondary school graduation, the graduation rate would have been 28 percent. |
percent of British 17 year olds and 12 percent of 18 year olds were attending secondary school full time in 1991. Students preparing for A-level exams achieve at high levels, but they represent a decided minority of their age cohorts. By contrast, French and Dutch youth have higher enrollment rates than American youth. For example, 86.4 percent of French and 90 percent of Dutch 17 year olds were in secondary school in 1991, but only 74.7 percent of American 17 year olds were. At age 18 enrollment in either secondary or tertiary education was 76 percent in France, 80 percent in the Netherlands, and 54 percent in the United States. Despite lower college attendance rates in France and the Netherlands, larger shares of 18 to 21 year olds in France (52.2 percent on a full-time equivalent [FTE] basis) and the Netherlands (56.4 percent) are enrolled in school (either secondary or tertiary) than in the United States (40.4 percent). Between ages 16 and 29, the average American spends 4.1 FTE years in school, British youth 2.3 years, French youth 4.6 years, and Dutch youth 4.9 years (OECD, 1993). These statistics contradict the widely held belief that the American education system, despite all its faults, at least achieves higher levels of participation than the European systems.
Not only are secondary school graduation standards higher elsewhere than in the United States, graduation rates are higher as well. In 1991 the graduation rate was 82.2 percent in the Netherlands, 75.8 percent in France, and 75.5 percent in the United States. The large proportions of 18 to 19 year olds attending secondary school in France and the Netherlands indicate how high graduation standards are made compatible with high graduation rates. Students having difficulty with the fast-paced curriculum do not drop out; rather, they repeat grades and thus gain extra time to prepare for the demanding external exams. Many participate in vocational programs and apprenticeships, which currently account for 54 percent of French and 70 percent of Dutch upper secondary students (OECD, 1993).
The benefit of earlier completion of secondary school in the United States is that large numbers of students enter tertiary education at a young age. However, some of the material covered during the first two years of college in the United States is covered in upper secondary school in France and the Netherlands. More bachelor's degrees are awarded in the United States, but some doubt that the B.A.s awarded by America's second-rank universities represent the same standard of achievement as comparable European degrees. Hard evidence on this issue is not available.
Teacher Quality and Compensation
The quality of the people recruited to teach is very important. A teacher's general academic ability and subject knowledge are the characteristics that most consistently predict student learning (Hanushek, 1971; Strauss and Sawyer, 1986; Ferguson, 1990; Ehrenberg and Brewer, 1993; Monk, 1992).
Secondary school teaching is not a prestige occupation in the United States,
and it apparently does not attract the same level of talent as in France and the Netherlands. Since university admission standards are higher in Europe, the university graduate pool from which European secondary school teachers are recruited is better educated on average than the college graduate pool out of which American teachers are recruited. Furthermore, American teachers are generally not the most talented members of the pool of college graduates. In 1977–1978 the mathematics Scholastic Aptitude Test (SAT) score of intended education majors was .38 standard deviations (SDs) below the overall average, 1 SD below engineering majors, and 1.2 SDs below physical sciences majors. The verbal SAT scores of intended education majors were .30 SDs below the overall average (NCES, 1992). In this respect, Britain is similar; entrants into programs to prepare primary school teachers have significantly lower A-level grades than average for university entrants (O'Leary, 1993).
In France, by contrast, secondary school teachers must do a double major in the two subjects for which they seek certification and then pass rigorous subject matter examinations. In 1991 only 31.3 percent of those who took the written exam for the Certificat d'Aptitude au Professorat de l'Enseignement du Secondaire, the most common of these examinations, passed. The best teaching jobs go to those who pass an even more rigorous examination, the Agregation Externe, which in 1991 had a pass rate of 17.7 percent (Ministere de l'Education Nationale et de la Culture, 1992a and 1992b). French and Dutch secondary school teachers tend to be recruited from the middle of a pool of graduates of tertiary education, which in turn is a more selected sample of the nation's population.
Furthermore, American teachers are often not expert in the fields they teach. Recent college graduates recruited into math or science teaching jobs spent only 30 percent of their college career taking science and mathematics courses. Since 46 percent had not taken a single calculus course, the prerequisite for most advanced mathematics courses, it appears that most of the math taken in college consisted of reviewing high school mathematics (NCES, 1993). The graduates of the best American universities typically do not become secondary school teachers because the pay and work conditions are relatively poor.
Compensation. The high academic standards for entry into upper secondary school teaching in France and the Netherlands are sustainable only if wages and work conditions are attractive. Data on the relative compensation of secondary school teachers are presented in rows 1 and 2 of Table 7.4. American upper secondary school teachers start at a wage that is 14 percent below that of the average worker, and after 15 years of experience they earn only 33 percent more. Starting salaries are equally low in England. However, the starting salaries in France are 6 percent above the average for all workers and in the Netherlands they are 39 percent higher. In France, England, and Scotland, upper secondary school teachers with 15 years of experience are paid 61 to 63 percent more than the average worker, and in the Netherlands they are paid 132 percent more than
TABLE 7.4 Teacher Compensation and Conditions of Work
|
France |
Netherlands |
England |
Scotland |
United States |
Compensation-Teacher/All Employeesc |
|
|
|
|
|
Upper secondary school teacher-starting |
1.06 |
1.39 |
.87 |
91 |
.86 |
Mid-career (15 yr) |
1.61 |
2.32 |
1.63 |
1.61 |
1.33 |
Lower secondary school teacher-starting |
.95 |
1.12 |
.87 |
.91 |
.86 |
Mid-career (15 yr) |
1.44 |
1.58 |
1.63 |
1.61 |
1.33 |
Primary school teacher-starting |
.93 |
.97 |
.87 |
.91 |
.84 |
Mid-career (15 yr) |
1.34 |
1.39 |
1.57 |
1.61 |
1.30 |
Teacher Class Contact Hours/Yeard |
|
|
|
|
|
Upper secondary school |
532 |
943 |
776 |
887 |
825 |
Lower secondary school |
706 |
943 |
776 |
887 |
748 |
Primary School |
875 |
1014 |
1013 |
950 |
1098 |
Class Sizee |
|
|
|
|
|
Upper secondary school |
29 |
24 |
16 |
15 |
25.6 |
Lower secondary school |
24 |
28 |
16 |
20 |
26.8 |
Primary School |
23 |
25 |
25 |
20 |
24.0 |
Secondary School Students/Teachersf |
14.0 |
15.9 |
14.7 |
14.7 |
15.5 |
Secondary School expenditure/student relative to GDP per capita |
28.1 |
24.7 |
28.0 |
28.0 |
29.4 |
Share of staff not classroom teachersg |
36% |
20% |
- |
- |
47% |
SOURCES: Nelson and O'Brien (1993), OECD (1993), Ministere de l'Education Nationale et de la Culture (1992a and b), and NCES (1992). a Compensation of secondary school teachers was calculated by multiplying their salary by the ratio of compensation to wages for manufacturing workers. This estimate of teacher compensation was then divided by the average compensation of all workers. The figure for French upper secondary school teachers is a weighted average of salaries for Agregé (20%) and others (80%). b Mean number of hours teaching a class per week times the mean number of weeks in the school year. Time devoted to preparation, in-service training, and to nonteaching activities is not included in this total. c Mean number of students in each class. d The ratio of the number of FTE pupils enrolled in public and private secondary schools to the number of FTE secondary school teachers. e Share of all staff employed in publicly funded elementary and secondary schools and ministries of education that are not classroom teachers. The nonteaching staff includes administrators at all levels, teachers aides, guidance counselors, librarians, nurses, custodial staff, food service workers, bus drivers, and clerical workers. The Dutch figure is for all three levels of schooling. The French figure is for secondary education only. The U.S. figure is for public elementary and secondary schools and does not include people working for state departments of education. In the U.S. teachers aides account for 8.8 percent of school staff. |
the average worker. For primary school teachers, by contrast, American pay levels are comparable to their Dutch and French counterparts (see row 6).
The lower pay in the United States is not compensation for more attractive work conditions (see rows 7–13 of Table 7.4). French upper secondary school teachers are in front of a classroom only 532 hours per year. Their American counterparts teach 825 hours per year. Teaching hours in England and Scotland are similar to U.S. levels, 776 and 886, respectively, but class sizes are substantially
smaller. Dutch upper secondary school teachers are the only group that clearly have heavier teaching loads than American teachers (Nelson and O'Brien, 1993).
When the salaries of college graduates are compared, those who enter teaching come out at the bottom. The starting salaries of U.S. mathematics and physical science majors who entered teaching were 42 percent below the salaries of those who obtained computer programming and system analyst jobs and 35 percent below the starting salaries of those obtaining jobs in mathematics or the physical sciences (NCES, 1993). University graduates who majored in a physical science earned 78 percent more and economics majors earned 92 percent more than education majors over the course of their working lives (Kominski and Sutterlin, 1992). Since Americans with university training in mathematics and science can earn much more outside teaching, those with talent in these areas are difficult to recruit into high school teaching. The result is that most American teachers of mathematics and science are less well prepared than their Northern European counterparts. This may help explain why American students lag French and Dutch students in mathematics and science but not reading. The fact that American primary school teachers are paid almost as much as French and Dutch teachers may also help explain why American 9 to 10 year olds compare favorably to their counterparts abroad.
There is a deeper question, however. Why are the academic standards for entry into upper secondary school teaching in the United States set so low? Why are salaries so low? These questions will be addressed later.
Overall Spending per Pupil
Data on pupil-teacher ratios and spending per pupil are presented in rows 13 and 14 of Table 7.4 Pupil-teacher ratios are quite similar in the five countries, as are the ratios of spending per pupil to per-capita gross domestic produce (GDP). Consequently, ''low" overall levels of spending on K-12 education are not the cause of the lag in U.S. student achievement.
Priority Given to Academics
If American spending per pupil is comparable to that in our four comparison countries, why are salary levels lower? What happens to the money saved by paying lower teacher salaries? It is used to hire additional nonteaching staff. Nonteachers account for nearly one-half of the employees in public education in the United States but only one-fifth in the Netherlands and 36 percent of secondary education employees in France (see bottom row of Table 7.4). These staff members perform services (such as bus transportation, sports activities, before- and after-school day care, counseling, and occupational training) that are provided by other governmental organizations or the private sector in some other nations. The money also pays for the more attractive buildings, sports facilities,
large school libraries, numerous computers, and colorful texts that are typical of American secondary schools. In part, this reflects the fact that in the United States books, computers, and buildings are cheaper relative to teachers of constant quality. U.S. spending patterns also reflect different goals. Academic achievement is the overarching—some would say the only—goal of French and Dutch secondary schools. In the United States, academic achievement must compete with other goals. American schools are also expected to foster self-esteem and provide counseling, supervised extracurricular activities, musical training, health services, community entertainment (such as interscholastic sports), and drivers' education—and do so in a racially integrated setting. These other goals require additional and different kinds of staff members. They may not be served by hiring teachers with a strong background in calculus or chemistry, so resources get diverted from paying the high salaries necessary to recruit teachers who are thoroughly educated in chemistry. Unlike France, selection into teaching is not based almost solely on competence in the subject matter.
The question remains, however, of why American school administrators give academic achievement lower priority than French and Dutch administrators do. This question will be taken up later.
Time Devoted to Instruction
Many studies have found learning to be strongly related to time on task (Wiley, 1976; Walberg, 1992). How do the five countries differ in the time that students spend in classrooms and doing homework? Table 7.5 reports the results of a variety of studies that compare time devoted to instruction. While estimates vary across studies, the pattern for secondary school students in the 1980s and 1990s is that French, Dutch, and Scottish students spent 5 to 15 percent more time in school than U.S. students. English students, by contrast, spent 6 to 9 percent less time in school than U.S. secondary school students.3
Differences in instruction time may explain some achievement differentials between countries, but they do not explain the generally poor showing of U.S. secondary school students in mathematics and science. While American students spend less total time in school, they get more mathematics and science instruction time than do French, Dutch, and Scottish students. Heavy European time commitments to foreign language study tend to crowd out mathematics and science instruction. In lower secondary school, British students study one foreign language and French and Dutch students generally study two. In America, by
3 |
Estimates of total time students in a country spend in school seem to depend on who is asked and how the question is worded. The data quality problem was dealt with by calculating an average across studies. The total instruction per year for each country was first expressed as a ratio to the U.S. level. Then a mean ratio was calculated by averaging the ratios from the studies that provided a comparison with the United States. Sources are given in the notes to Table 7.5. |
TABLE 7.5 Student Time—Instruction and Homework
|
France |
Netherlands |
England |
Scotland |
United States |
Total hours of instruction/year |
|
|
|
|
|
Primary school—1971 |
918 |
1040 |
900 |
1040 |
900 |
Grade 5 in 1982 |
— |
— |
984 |
— |
1070 |
Grade 4 in 1991 |
840 |
975 |
— |
— |
954 |
Secondary School—1971 |
775 |
1120 |
900 |
1080 |
900 |
Grade 9 in 1982 |
— |
1007 |
1025 |
— |
1141 |
Grade 8 in 1982 |
1187 |
1000 |
896 |
1067 |
1008 |
Grade 9 in 1991 |
1030 |
1092 |
— |
— |
792 |
Age 13 in 1991 |
1073 |
— |
960 |
1031 |
1003 |
Hours of Homework in All Subjects |
|
|
|
|
|
Hours/week—grade 9 in 1982 |
— |
8.4 |
6.0 |
— |
9.6 |
Hours/week—grade 8 in 1982 |
8 |
5 |
5 |
3 |
5 |
Hours/week—grade 12 math in 1982 |
— |
— |
— |
— |
9 |
Hours/week—grade 12 science in 1982 |
— |
— |
11.5 |
— |
9.8 |
Hours/week—grade 4 in 1991 |
0.53 |
0.13 |
— |
— |
1.89 |
Homework GT 2 hours/day, age 13 in 1991 |
55% |
— |
30% |
15% |
30% |
Hours worked on language arts—grade 4 |
9 |
7 |
— |
— |
11 |
No. language arts homework assignment/week, grade 9 in 1991 |
1.6 |
.4 |
— |
— |
2.3 |
Time Devoted to Mathematics |
|
|
|
|
|
Math share—grade 8 in 1982 |
12% |
10% |
13% |
14% |
14% |
Hours/week math instruction—age 13 in 1991 |
3.83 |
— |
3.17 |
3.50 |
3.80 |
Hours/week math homework—age 13 in 1991 |
1.93 |
— |
1.27 |
1.00 |
1.52 |
Hours/week math homework—grade 8 in 1982 |
4.0 |
2.0 |
1.0 |
2.0 |
3.0 |
Time Devoted to Science |
|
|
|
|
|
Grade 5 in 1971 |
8% |
2% |
3% |
3% |
7% |
Grade 5 in 1982 |
— |
— |
4% |
— |
10% |
Grade 9 in 1971 |
8% |
7% |
8% |
5% |
10% |
Grade 9 in 1982 |
— |
25% |
10% |
— |
20% |
Hours/week science instruction, age 13 in 1991 |
2.90 |
— |
3.23 |
3.00 |
3.88 |
Hours/week science homework, age 13 in 1991 |
0.68 |
— |
0.97 |
0.65 |
_1271.06 |
SOURCES: Passow et al. (1976), Postlethwaite and Wiley (1992), Robitaille and Garden (1989), IAEP (1992a, b), Lundberg and Linnakyla (1993). |
contrast, few lower secondary school students study a foreign language and, by high school graduation, students have taken an average of only 1.46 years of a foreign language (NCES, 1992).
European students learn mathematics and science more thoroughly than American students do even when they spend less time on it. For example, in the IAEP study, mathematics instruction time was the same in France and the United States, yet French students knew about 1.47 U.S. grade-level equivalents more than American students. In science, by contrast, instruction time was one hour
per week less in France, yet Americans still trailed French students by about one-third of a U.S. grade-level equivalent. Why does an hour of instruction in French and Dutch classrooms produce more learning than in American classrooms? Could heavier homework assignments be the explanation?
Cooper's (1989) meta-analysis of randomized experimental studies found that students assigned homework scored about one-half of a standard deviation higher on posttests than students not receiving homework assignments. The impact of homework on the rate at which middle school students learn also was significant, although somewhat smaller. Non-experimental studies using IAEEA and IAEP data come to similar conclusions.
French lower secondary school students spent more time doing mathematics homework and homework of all types (see Table 7.6). For example, 55 percent of their 13-year-olds reported doing over two hours of homework a night, compared to 30 percent in the United States and England and only 15 percent in Scotland.4 This is consistent with their lead in mathematics achievement. In science, however, there is no evidence that Dutch and French students had more homework than American students. Furthermore, English and Scottish lower secondary school students do less homework and have less instruction time in mathematics and science than American students but still outperform them.
Engagement—Effort per Unit of Scheduled Time
Classroom observation studies reveal that American students actively engage in learning activities for only about half the time they are scheduled to be in a classroom. A study of schools in Chicago found that public schools with high-achieving students averaged about 75 percent of class time for actual instruction; for schools with low-achieving students, the average was 51 percent of class time (Frederick, 1977). Overall, Frederick et al. (1979) estimated 46.5 percent of potential learning time is lost due to absences, lateness, and inattention.
Just as important as the amount of time spent participating in a learning activity is the intensity of the student's involvement in the process. At the completion of his study of American high schools, Sizer (1984) characterized students as "all too often docile, compliant, and without initiative" (p. 54). Goodlad (1983) describes "a general picture of considerable passivity among students" (p. 113). The high school teachers surveyed by Goodlad ranked "lack of student interest" as the most important problem in education. In a Longitudi
nal Survey of American Youth (1989) 62 percent of 10th graders agreed with the statement "I don't like to do any more school work than I have to."
Formal studies comparing ratios of on-task time to scheduled time are not available for European countries. Nevertheless, people who have visited classrooms in France or the Netherlands and the United States report that European teachers are less likely to talk about extraneous matters and that European students are more likely to pay attention and do their assignments. My own school visits in France and the Netherlands generated similar impressions.
Summary
Four of the seven proposed explanations for American students trailing French, British, and Dutch students in math and science can be ruled out: diversity, restricted access, spending per pupil, and time for instruction. Three hypotheses survive the first round of tests: lower-quality teachers, lower priority attached to academic goals, and lower levels of student engagement. With only five data points, no further narrowing of the list of hypothesized proximate causes is possible. Now let us look behind these proximate causes for ultimate causes. Why does an hour of instruction and homework time apparently have larger learning effects in England, France, and the Netherlands than in America? Why do French and Dutch secondary school mathematics and science teachers apparently expect more of their students than American teachers do? The next section of this chapter proposes some tentative system-level answers to these questions. The purpose is to show that a very simple application of economic theory can provide a plausible explanation for the large system-level differences in goals and learning efficiency cited above.
Signaling as the Ultimate Cause: External Examinations as Standard Setters
In a 1990 paper I proposed the following answer to these questions:
The fundamental cause of the low effort level of American students, parents, and voters in school elections is the absence of good signals of effort and learning in high school and a consequent lack of rewards for effort and learning. … In most other advanced countries mastery of the curriculum taught in high school is assessed by … examinations which are set and graded at the national or regional level. Grades on these exams signal the student's achievement to colleges and employers and influence the jobs that graduates get and the universities and programs to which they are admitted. How well the graduating seniors do on these exams influences the reputation of the school and in some countries the number of students applying for admission to the school. In the United States, by contrast, students take aptitude tests that are not intended to assess the learning that has occurred in most of the classes taken in high school.
The primary signals of academic achievement are grades and rank in class—criteria which assess achievement relative to other students in the school or classroom, not relative to an external standard. (Bishop, 1990, p. 3.)
Costrell (1994a, b) formally modeled the setting of educational standards and concluded that decentralized standards setting (i.e., teacher grading or school graduation requirements) results in lower standards, lower achievement, and lower social welfare than does more centralized standards setting (state or national achievement exams). He also concluded that "the case for perfect information [making scores on external examinations available rather than just whether an individual passed or failed] would appear to be strong, if not airtight: for most plausible degrees of heterogeneity, egalitarianism, and pooling under decentralization, perfect information not only raises GDP, but also social welfare" (1994a, p. 970).
Of the 50 states, only New York has a system of curriculum-based achievement exams that affect individual student grades and are taken by large shares, about one-half, of high school students. This or something else unique to New York state appears to have raised achievement levels. Graham and Husted (1993) discovered this fact when they examined the determinants of mean SAT test scores in the 37 states with reasonably large test-taking populations. Controlling for the proportion of high school seniors taking the SAT and the race, gender, parental income, and parental education of the test takers, they found that New York state had the highest adjusted mean SAT scores. They did not, however, test the statistical significance of the New York state effect and used an unusual log-log specification.
Are their findings robust to changes in specification? How large is the difference between New York and the rest of the nation? Is the differential statistically significant? Table 7.6 presents the results of a regression predicting 1991 mean SAT math and verbal scores for the 37 states for which data are available. With the exception of the dummy variable for New York state, all right-hand-side variables are proportions—generally the share of the test-taking population with the characteristic described. Clearly, New Yorkers do significantly better on the SAT, particularly the math portion, than do students of the same race and social background living in other states.5 For individuals the
TABLE 7.6 Determinants of Mean SAT Scores for States
|
New York State |
Participation Rate |
Parents AA-BA+ |
Private School |
Proportion Black |
Large School |
3 + Math Courses |
3 + English Courses |
R2 RMSE |
Total |
46a |
68a |
370b |
60 |
-135b |
-44a |
85 |
-36 |
.93 |
SAT |
(2.7) |
(2.6) |
(6.4) |
(1.6) |
(3.2) |
(1.8) |
(1.3) |
(.3) |
14.8 |
Mean |
.027 |
.414 |
.581 |
.207 |
.078 |
.129 |
.617 |
.797 |
Tot SAT |
|
|
|
|
|
|
|
|
|
55 |
StanDev |
.164 |
.240 |
.097 |
.082 |
.064 |
.113 |
.067 |
.038 |
925 |
a significant at 5% level. b significant at 1% level. |
summed SAT verbal and math scores have a standard deviation of approximately 200 points. Consequently, New York state's SAT mean is about 0.2 SDs or about 0.75 grade-level equivalents higher than the regression's prediction.
This occurred despite that fact that Regents exams involve very modest stakes. Exam grades account for less than one-half of final course grades and influence only the type of diploma received. A passing score on a Regents exam is not necessary for admission to community colleges and employers often ignore exam results when they make hiring decisions.
How do the examination systems of our four comparison countries work? In 1992, 71 percent of French youth took a Baccalaureat (Bac) exam. Fifty-one percent of the age group passed. Thirty-eight percent of the baccalaureates awarded were Bac Technologique or, in vocational lines, the Bac Professionel (Ministere de L'Education Nationale, 1992a and b). This was a major accomplishment, for Bac exams are set to a very high standard. The three-year lycee programs that prepare 43 percent of the age cohort for the Bac General are quite rigorous. Bac exams in mathematics, history/geography, and French are set and marked by 23 regional academies. School-based assessments are used for other subjects (Madeus and Kellaghan, 1991). The Bac exams taken in any one area of concentration are comparable to the advanced placement exams taken by American students seeking college credit for high school work. Cornell University, for example, generally awards advanced placement credit to recipients of the Baccalaureat General.
In France the payoff to higher education is high, so access to a university is highly prized. A Bac is necessary for university admission and the line pursued and the mentions obtained on the exam influence which university program a student can enter.6 About 10 percent of those obtaining a Bac General enter special programs that prepare them for the exam regulating admission to the elite Grandes Ecoles. The job market also rewards young people who have passed the Bac. There are alternative lower-level qualifications for employment such as the Brevet d'Enseignment Professionnel (BEP) and the Certificat d'Aptitude Professionelle (CAP), but the Baccalaureat confers greater access to preferred jobs. In 1987, unemployment rates for 15 to 24 year olds were in France 37 percent for those without a diploma, 22 percent for those with CAPs or BEPs, 18 percent for those with a Bac, and 10 percent for university graduates (Ministere de l'Education Nationale, 1992b).
Dutch university graduates ages 45 to 64 earn 65 percent more than secondary school graduates (OECD, 1992, 1993), so access to higher education is highly prized in the Netherlands as well. Examinations set by the Ministry of Education influence access to postsecondary education, so the high achievement of Dutch
students in mathematics and science can be explained in the same way.7 In both France and the Netherlands, questions and answers are published in newspapers and available on video text. The published exams signal the standards that students and teachers must aim for.
Nine-tenths of English youth now take the General Certificate of Secondary Education (GCSE) exam at the end of eleventh grade, and an increasing number take A-level exams two years later. Scotland also has a system of external examinations. For the United Kingdom as a whole, the ratio of the number of school leavers passing at least one A-level exam (or the Scottish equivalent) to the number of 19 year olds was 23 percent in 1991 (Government Statistical Service, 1993). Completing an A-level qualification lowers unemployment rates for 25 to 34 year olds from 16.9 to 6.9 percent and graduating from university lowers it further to 4.3 percent. University graduates earn 66 percent more than secondary school graduates at ages 45 to 64 (OECD, 1992). Performance on the GCSE and A-level exams and the equivalent Scottish exams determines whether a student can continue his or her schooling and which university and program he or she can enter. Grades on the GCSE and A-level exams are included on resumes and requested on job applications, so employment opportunities depend on school results as well (Raffe, 1984).
In the United States, by contrast, admission to the best colleges depends on teacher assessments of relative performance—rank in class and grades—and aptitude tests that are not keyed to the courses taken in secondary school.
External assessments of achievement that directly affect access to preferred educational and job outcomes clearly increase students' rewards for studying. They also change the structure of rewards for learning and, therefore, the incentive environment of students, teachers, and administrators. I will argue that the structure of rewards for study is at least as important as their size. These issues will be discussed under seven headings:
- Peer group norms
- Teacher incentives
- Administrator incentives
- Competition among upper secondary schools
- Standards of the external exams
- Redoublement as mastery learning and an incentive to study
- Choice of specialization as goal setting
Peer Group Norms
In the United States peer groups often try to discourage academic effort. No adolescent wants to be considered a "nerd, geek, or grade grubber." Nor do blacks want to be accused of "acting white." That, however, is what happens in most classrooms to students who study hard. Because the school's signals of achievement assess performance relative to fellow students through grades and class rank, not relative to an external standard, students have a personal stake in persuading each other not to study.
An important reason for peer pressure against studying is that pursuing academic success forces students into a zero-sum competition with their classmates. Their achievement is not being measured against an absolute external standard. In contrast to scout merit badges, for example, where recognition is given for achieving a fixed standard of competence, the school's measures of achievement assess performance relative to fellow students through grades and class rank. A student who does well on exams makes it more difficult for other members of the class to get an A or to be ranked at the top of the graduating class. Since devoting time to studying for an exam is costly, the welfare of an entire class is maximized if no one studies for exams that are graded on a curve. The cooperative solution is "no one studies more than the minimum." Participants are generally able to tell who has broken the "minimize studying" code and reward those who conform and punish those who do not. Side payments and punishments are made in a currency of friendship, respect, and ridicule that is not limited in supply. For most black students the benefits that might result from studying are less important than the very certain costs of being considered a "geek" or "acting white," so most students abide by the ''minimize studying," "don't raise your hand too much" norm.
The peer norms that result are: "It's OK to be smart. You can't help that. But it is definitely not OK to study hard to get a good grade." This is illustrated by the following story related by a Cornell undergraduate:Erroneously I was lumped into the brains genus by others at [high] school just because of the classes I was in. This really irked me; not only was I not an athlete but I was also thought of as one of those "brain geek." Being a brain really did have a stigma attached to it. Sometimes during a free period I would sit and listen to all the brains talk about how much they hated school work and how they never studied and I had to bite my lip to keep from laughing out loud. I knew they were lying, and they knew they were lying too. I think that a lot of brains hung around together only because their fear of social isolation was greater than their petty rivalries. I think that my two friends who were brains liked me because I was almost on their level but I was not competitive.
Note how those who broke the "minimize studying" norm tried to hide that fact from their classmates. They did not espouse an alternative "learning is fun and important" norm.
The costs and benefits of studying vary across students because interest in any given subject varies, ability varies, parental pressure varies, and rewards vary. This heterogeneity means that some students break the "minimize studying" norm. When they are a small minority, they cannot avoid feeling denigrated by classmates. In the top track and at schools where many students aspire to attend competitive colleges, they are numerous enough to create a subculture of their own, with its own norms denigrating those who do poorly on tests or who disrupt classroom activities. This is the structural basis of the "brains" and "preppie" cliques found in most American high schools. Most high school students, however, are in cliques that denigrate studying. At some school awards ceremonies, some in the crowd jeer as students are called to come up to receive awards (Suskind, 1994).
Peer pressure was discussed in my interviews of school staff members and students in England, the Netherlands, and France. The French educators I spoke to reported that peer pressure not to study occurred sometimes but only in some lower secondary school classes, not at the lycee serving upper-middle-class students that I visited. In lower secondary schools the pressure appeared mild by American standards. In upper secondary schools, particularly in the math-science line, the peer pressure was to excel. Discussions with Dutch and English students and educators produced similar observations.
Teacher Incentives
Most American secondary school teachers do not feel individually accountable for the learning of their students. This lack of accountability for learning stems from (1) the rarity of examinations that assess student achievement in particular subjects relative to an external standard and (2) the fact that most secondary school students receive instruction in a given subject from many teachers. Only coaches, band conductors, and teachers of advanced placement classes are exceptions. They teach in environments where student achievement is visible to parents and colleagues and, as a result, feel accountable for outcomes.
In France and the Netherlands, by contrast, upper secondary students are grouped in small classes, take most subjects together, and generally are together for two or more years. Fewer than three teachers share responsibility for preparing each class for the external exams. In the Netherlands, where schools are small, many subjects are taught by only one teacher. Since important rewards accrue to those who pass or do well on exams, everyone takes them seriously. The number of students taking and passing each exam is public knowledge within the school and among parents. Exam results influence teachers' reputations. Responding to such informal pressures, upper secondary school teachers strive to prepare their students for the external exams.
American teachers also are expected to ensure that most of their students pass, but they are free to accomplish this goal by lowering the passing standard.
Teachers who set expectations that are too high can get into trouble. For example, Adele Jones, an algebra teacher in Georgetown, Delaware, was fired because she failed too many of her students—42 percent one year and 27 percent the next. When students started picketing the school carrying "hastily scrawled signs with such slogans as 'I Failed Ms. Jones's class and It Was My Fault' and 'Just Because a Student Is Failing Doesn't Mean the Teacher Is'" (Bradley, 1993) the national news media took notice. The principal of the school justified his decision as follows:
I have made it very clear that one of my goals is to decrease the failure rate, to make sure the kids feel good about learning, stay in class, stay in school and do well.… Math is just a big body of knowledge; what is Algebra II across the nation anyway?" he asks. When he taught band, he adds, he certainly didn't expect kids to finish the year as musicians—but he did want them to know more about music than … before.… The talk about preparing students for college struck him as "ludicrous." Instead the goal should be to keep students studying math. (Bradley, 1993, pp. 19, 20)
Senior Norman Kennedy said, however, that the students who flunked Ms. Jones's class "were sleeping. They don't want to learn. They goof off, and they talk." At the hearing Walter Hall, Jr., a student who had flunked the course, testified:
I guess some of it could be attributed to a lack of study, because I wasn't really like into the books hour after hour. But in the rest of my classes, I was doing fairly well, and it was only testing that gave me a problem." He added that his parents had wondered how he could be getting such good grades in most classes without studying. (Bradley, 1993, p. 20)
A survey of teachers by Peter D. Hart Research Associates (1994) found that 30 percent reported "feeling pressure to … give higher grades than students' work deserves." Forty-six percent reported pressure to "pass students on to the next grade who are not ready." Thirty percent reported pressure to "reduce the difficulty and amount of work you assign."
Ms. Jones is unusual; most teachers realize that they must limit their failure rate. More commonly, the struggle over expectations plays out in the privacy of the classroom. Sizer's (1984) description of Ms. Shiffe's biology class, illustrates what sometimes happens:
She wanted the students to know these names. They did not want to know them and were not going to learn them. Apparently no outside threat—flunking, for example—affected the students. Shiffe did her thing, the students chattered on, even in the presence of a visitor.… Their common front of uninterest probably made examinations moot. Shiffe could not flunk them all, and, if their performance was uniformly shoddy, she would have to pass them all. Her desperation was as obvious as the students' cruelty toward her. (pp. 157–158)
Some exceptional teachers are able, through the force of their personalities, to induce students to undertake tough learning tasks. But for all too many aca-
demic demands are compromised because the bulk of the class sees no need to accept them as reasonable and legitimate.
Administrator Incentives
External assessment changes the incentives faced by school administrators. In the United States locally elected school boards and the administrators they hire make the thousands of decisions that determine academic expectations and program quality. When there is no external assessment of academic achievement, students and their parents benefit little from administrative decisions that opt for higher standards, more qualified teachers, or a heavier student workload. The immediate consequences of such decisions—higher taxes, more homework, having to repeat courses, lower grade point averages (GPAs), less time for fun courses, a greater risk of being denied a diploma—are all negative. When student learning is not assessed externally, the positive effects of choosing academic rigor are negligible and postponed. Since college admission decisions are based on rank in class, GPAs, and aptitude tests and not externally assessed achievement in high school courses, upgraded standards will not improve the college admissions prospects of a secondary school's graduates. Graduates will do better in difficult college courses and will be more likely to get a degree, but that benefit is uncertain and far in the future. Maybe over time the school's reputation and, with it, the admissions prospects of graduates will improve because the current graduates are more successful at local colleges. That, however, is an even more uncertain and delayed result.
Few American employers pay attention to a student's achievement in high school or the school's reputation when they make hiring selections (Bishop, 1993; Hollenbeck and Smith, 1984). Those who do pay attention to school achievement use such indicators of relative performance as GPA and class rank rather than results on an external exam as a hiring criterion. Consequently, higher standards do not benefit students as a group, so parents as a group have little incentive to lobby strongly for higher teacher salaries, higher standards, and higher school taxes. Employers who recruit from a local high school are often the only group with a real interest in general increases in achievement. Since, however, they pay a disproportionate share of school taxes, they tend to support only policy options that do not cost additional money.
By contrast, in many European countries the record of each school in the external examination—the number of students who pass or get high grades—is published in local and national newspapers. Recent reforms in England and Scotland, for example, have resulted in schools publishing annual reports that contain the grades received by last year's students in each subject tested. These reports are sent to parents of current and prospective students. The school league tables have important effects on school reputations. Administrators seeking to
strengthen their school's reputation are thus induced to give teaching effectiveness, as assessed by the external exam, first priority.
Competition Among Upper Secondary Schools
For generations French and Dutch upper secondary schools have faced a competitive environment that is similar in many ways to the one faced by American colleges and universities. Funding has been on a per-student basis, so schools experiencing an increase in applications have had an incentive to expand up to the capacity of their physical plant. Schools with strong reputations get more applications than they can accept and are, in effect, rewarded by being allowed to admit the "best" from their pool of applicants.
In the United States access to quality teaching and supportive peers depends on parental ability to buy or rent a home in a suburb with excellent schools. In France and the Netherlands access to the top upper secondary schools depends primarily on achievement in lower secondary school. This means that parents who want their child to attend the best upper secondary schools must make sure their child does well in lower secondary school.
The Netherlands has three types of general secondary schools—the VWO (pre-university, secondary education institution-most difficult), HAVO (senior general secondary education institution-next most difficult), and MAVO (junior general secondary education institution-least difficult)—and a system of lower vocational schools—LBO/LEAOs (junior secondary vocational education institutions) and KVBOs (agricultural junior secondary vocational education institution)—that prepare students for both occupation-specific and general education exams. The first-year curriculum is supposed to be the same in all schools, so that students can transfer between schools if necessary. In succeeding years, the curricula and rigor diverge. Rigor and workloads are greatest at the six-year VWOs, somewhat less demanding at the five-year HAVOs, and still less demanding at the four-year MAVOs. These schools also differ in the foreign languages offered and the standard to which they are taught. The LBOs devote considerable time to occupation-specific curricula, so less time is available for general studies. Advice to parents about which type of school is appropriate for their child is based on the pupil's record in primary school and in some cases standardized tests (Nijhof and Streumer, 1988). Parents have the right, however, to select the type of school and which school of that type their child will enter. In addition, there are three parallel systems of education—a locally administered public system, a Catholic system, and a Protestant system—so parents have a great deal of choice.
About a decade ago English and Scottish parents were given the right to send their children to schools outside their normal attendance areas. Two years after choice became operational in Scotland, 9 percent of pupils entering secondary school nationally (11 to 14 percent in urban areas) attended a school outside their
cachement area (Adler and Raab, 1988). Scottish parents who made this choice appeared to be behaving rationally, for they tended to choose schools that were more effective than the school in their home area. An analysis of school choice in the Fife Education Authority found that the schools chosen by those leaving their cachement area had better examination results than would have been predicted given the pupil's primary school test scores and family background and the average socioeconomic status of the pupils at the school.8 Consequently, the free choice of schools that prevails in our four European nations generates a competitive pressure on schools to excel that has no counterpart in the United States outside cities with magnet schools.
Standards of the External Exam
External examinations at the end of secondary school are probably necessary if high achievement levels are to be attained, but they are not sufficient. Effects will be small if the exams are easy, are taken by only a small minority of students, or do not generate substantial rewards for successful students. British youth have lower achievement levels than French and Dutch youth. One possible explanation for this is that the passing standard of the GCSE is lower than for the Bac and the Dutch exams, and the more difficult A-levels are taken by only a small minority.
High passing standards on external exams are clearly associated with high achievement levels. Does this reflect a cause-and-effect relationship? Yes, but causation runs both ways. High passing standards on medium-and high-stakes exams are politically sustainable only when most students taking the exam are able to meet or surpass the standard. At present, the median pupil in Britain is not expected to learn the entire multiplication table up to 10 × 10 until age 11. If the GCSE mathematics exams were made more demanding without strengthening mathematics teaching, failure rates might rise to politically unacceptable levels.
Does the passing standard also influence student effort? Yes. In data for high school and beyond, those taking more rigorous courses learned a good deal more between their sophomore and senior years, even though their GPAs suffered as a result (Gamoran and Barends, 1987). Kulik and Kulik's meta-analysis (1984) of the educational literature found that students randomly assigned to skip
a grade or to a compressed and accelerated curriculum scored 75 percent of a standard deviation higher on tests (a few years later) than the matched non-accelerated students. Repeating a grade effectively lowers learning goals and reduces the retained child's achievement a few years later by about 30 percent of a standard deviation (Holmes, 1989).
Over 100 experimental studies have been conducted of the effect of goal difficulty on various kinds of achievement. The effects are quite large. On highly complex tasks such as school and college course work, specific hard goals raised achievement by 47 percent of a standard deviation (Wood et al., 1987). In the laboratory and field settings used by psychologists conducting this research, the subjects generally accepted the goal set for them by the researcher. Achievement went up, but the probability of failing to reach the goal rose as well. In most studies more than two-thirds of those in the "hard goal" condition failed to achieve their goal (Locke, 1968). Most studies examined behavior over relatively short periods of time. One would imagine, however, that if such experiments lasted a couple of years, those who consistently failed to achieve their goal might lower their goals or give up altogether.
Stedry (1960) found that when subjects who had already set their own goals were assigned even higher ones by the study director, they rejected the assigned goal and achievement did not rise. This appears to be what happens in American secondary schools. Most students reject the goals that teachers set because the rewards for success are small. Others reject them because they appear to be unattainable.
How do European education systems induce students in upper secondary schools to set difficult learning goals and work toward them? They do not, as some have proposed for the United States, set a single high yea-nay standard that everyone is expected to meet. Young people are too different from each other for such a policy to work.9 When exams are graded pass-fail and the same passing standard applies to all, many students are able to pass the standard without exertion and will, therefore, not be stimulated to improve by the need to pass the exam.10 Many other students will think they are now so far behind and the effort required to achieve the standard so great that the costs of the effort are larger than the possible reward. They will reject the goal of meeting the standard. When the variance of performance is large, only a few students will find the reward attached to a single absolute passing standard an incentive to study (Kang, 1985).
External exams need to signal the level of a student's achievement, not just whether the exam was passed. Dutch external exams are graded on a scale of 1 to 10. Excellence on the Baccalaureat exams results in the award of a Mention Tres Bien, a Mention Bien, or an Mention Assez Bien. Once information on performance levels becomes available, employers and institutions of higher education will tend to base their selection decisions on it. Graduates with the strongest exam results have options not available to those with weak results, and the outcome is a system of graduated rewards. When the variance of achievement is high, incentives for effort are stronger on average under a graduated rewards system than under a single large reward attached to achieving a fixed standard (Kang, 1985).
The English GCSE and Scottish "Lowers" Examinations are taken by 90 percent of 16 year olds. As recommended by Kang's model, they generate substantial and graduated rewards for learning what appears on the exams. Indeed, the rewards for doing particularly well on these external exams appear to be larger than those in the Netherlands. 11 Why then are English and Scottish 13 year olds assigned less homework than their American and Dutch counterparts? Why is their achievement in mathematics and science at age 13 significantly lower than in the Netherlands? As the time for the exam approaches in Britain, teacher demands and student effort increase substantially. At age 13, however, standards are low. Why do the backwash effects of the secondary school graduation exams extend further back in the pupil's schooling in the Netherlands and France than in Britain?
Redoublement as Mastery Learning and an Incentive to Study
One explanation for low British standards for 10 to 13 year olds is the lack of immediate rewards for doing well in classes. The external exams are three to six years away. Students are promoted to the next grade no matter how well they do in the previous grade. Those who fall behind inevitably slow the pace of the class in succeeding years. Primary school teachers do not feel accountable for how well students do on exams taken after four years of attendance at a secondary school. Secondary schools tend to be large, and the teachers who handle the first-year students lack a sense of accountability for performance on exams that are more than three years in the future.
The situation is very different in France and the Netherlands. Pupils who fail more than one of their courses are generally required to redouble or repeat the grade. In 1990 Dutch redoublement rates were 7.5 percent per year in academic lower secondary schools, 5.1 percent per year in LBOs (junior secondary vocational education institution), the vocational lower secondary schools, and 13.3 percent per year in academic upper secondary schools (Central Bureau Voor De Statistiek, 1993). French rates of redoublement ranged from 6.8 and 11.0 percent per year during the four years of general lower secondary education, 12.1 to 18.4 percent per year in the three-year academic upper secondary schools, and 8.4 percent per year in the first two years of vocational upper secondary schools (Ministere de l'Education Nationale et de la Culture, 1992a). According to Lewis (1985), the "basic motivation is to help the child himself, to ensure that the pupil is sufficiently well prepared so that he may fully benefit from work at a more demanding level" (p. 5). For French teachers, redoublement is a form of mastery learning, a way of allowing some students extra time to achieve very demanding learning goals. Consequently, at age 19, 31.6 percent of French and 41.5 percent of Dutch youth are still in secondary school, compared to 3.4 percent in Britain and 5 percent in the United States.
Redoublement is not something that is inflicted only on children from lower-class backgrounds. Often high aspirations can be achieved only by redoublement. The two Dutch professors with grown children with whom I have discussed this matter both had a child who was required to repeat a grade. In France selective upper secondary schools serving upper-middle-class communities have grade-repeating rates that are nearly as high as schools serving lower-income communities. For example, Lycee Charlemagne, an upper secondary school serving one of the richest neighborhoods in Paris, asked 14 percent of its entering class to repeat the year in 1992.
For French and Dutch teenagers the threat of having to repeat a grade is a strong incentive to study. When I asked how the students who must redouble feel about it, I was told that they feel "dishonored." Since redoublement is a public event, parents also feel stigmatized, so they have an incentive to see that their child studies hard. In the Netherlands, students struggling with the fast-paced VWO or HAVO curricula are often given a choice: either repeat the year or transfer to a less demanding school. At the VWO I visited in the Netherlands, one-third of the entering class transfers to a HAVO or a less demanding VWO before the beginning of the third year. VWOs offer a fast-paced six-year university preparation program. Parents who want their child to enter a VWO are generally accommodated even when primary school teachers advise against it. The child's performance in school determines whether the parents' aspirations are realized or whether a transfer to a less demanding type of school is necessary. Being forced to transfer to an HAVO or MAVO does not foreclose university attendance. With good grades at the end of the five-year HAVO program, a student can transfer to a VWO, complete the final two years, and then enter a
university. In addition, numerous vocationally oriented higher education options are open to HAVO and MAVO graduates and transfers to a university are feasible with good grades.
While other routes to a university education are possible, pupils who choose the fast track in seventh grade, a VWO, do not want to be forced "to get off the train." Students in the Netherlands and France are formed into classes that take most subjects together and remain intact for two years and sometimes longer. Friendships tend to develop within this class. When I asked a Dutch student who, despite long hours of study, had been required to repeat a grade, why she had studied so hard, she responded, "I wanted to stay with my class!" Students do not want to have to repeat the grade because it threatens to sever the friendships they have made in class. Apparently, trying to keep up academically or accepting the academic goals of the school is viewed positively by peers because it is an expression of commitment to the group. Those who refuse to study are apparently seen as rejecting the group. In these two countries peer pressure seems to encourage lagging students to study, not discourage them as in the United States. 12
Choice of Specialization as Goal Setting
All education systems give upper secondary students and their parents the right to select a specialty and the right to choose the rigor and difficulty level of either the school, the academic program, or specific courses.
In France four academic lines—literature and languages (A), economics and social sciences (B), mathematics and physical sciences (C), and biology (D)—have roughly equal numbers of students and together account for most of the Baccalaureat Generales awarded. The mathematics-physics-chemistry line (C) is the most difficult, carries the greatest prestige, and gives one the best chance of being admitted to a preparatory school for one of the elite Grandes Ecoles. Admission to the C line within a lycee is generally highly competitive. The Netherlands has a similar though less elaborate system of specialization within general
upper secondary education. As in France, the math-science line has the reputation of being the most difficult.
In France and the Netherlands, picking one's school and specialization effectively sets a specific learning goal. The prevalence of grade repeating and transfers to easier schools suggests that most students and parents initially set very difficult goals. The goal-setting literature tells us that working toward a specific and difficult goal leads to greater effort and performance than being told to ''do your best" or setting easy goals. Thus, the continental European pattern of setting highly ambitious goals maximizes average achievement levels even while it increases the number of students who fail to achieve the goal they initially set. Why do French and Dutch parents select secondary schools and programs that are so challenging that many must repeat grades to keep up or transfer into easier programs and schools? There are three reasons. First, the goal selected is visible to parents, relatives, and neighbors and going for difficult goals confers prestige. Second, achieving difficult learning goals is rewarded by admission to preferred universities and fields of study and access to better jobs. Finally, the choice is generally made by the parent, not the child. Parents are better informed about the long-term benefits of achieving difficult goals, and their own prestige rises when their child attends a selective school or pursues a difficult line of study. Parents may view the extra studying necessary in a rigorous specialty as a plus not a minus.
In America, by contrast, selecting difficult goals generates much weaker rewards. Everyone in the neighborhood attends the same school. Students select individual courses, not programs or schools. Subjects are taught at vastly different levels, but the rigor of the courses is not well signaled to parents, relatives, neighbors, employers, or colleges. Admissions staff at selective colleges learn how to read the transcripts of high schools they recruit from and evaluate grades in that light. However, many colleges have, historically, not factored the rigor of high school courses into their admissions decisions. Almost no employers do. Consequently, most students not aspiring to attend a selective college avoid rigorous courses and demanding teachers. As one student put it:
My counselor wanted me to take Regents history and I did for a while. But it was pretty hard and the teacher moved fast. I switched to the other history and I'm getting better grades. So my average will be better for college. Unless you are going to a college in the state, it doesn't really matter whether you get a Regent's diploma. (Ward, 1994, p. 1)
Another student who had avoided the harder courses even though she was sure she could do the work explained her decision with, "Why should I do it [the extra work] if I don't have to?" (Ward, 1994). Some students, the minority who want to attend selective colleges, sign up for demanding courses. Most students choose courses that have the reputation of being fun and not requiring much work to get a good grade. Teachers know this and adjust their style of teaching, assignments, and grading standards with an eye to maintaining enrollment levels.
Summary and Lessons
In the Netherlands and France, learning in secondary school is assessed by difficult subject-specific external examinations, and doing well on the exams generates large rewards for the student. The reputations of teachers and schools are affected by student achievement on the exams. Parents base their selection of the upper secondary school their child will attend and which academic or vocational program he or she will pursue, in part, on these reputations. Parents tend to set difficult goals for their children, so most students are placed in programs of study that for them are very demanding. Students are grouped into classes that take all their subjects together, remain intact for two years or more, and become the student's circle of friends. Students who are not progressing at the rate necessary to succeed on the external exam are asked to either switch to an easier curriculum or repeat the year. Students do not want to be forced to sever the friendships they have developed in their class, so they are strongly motivated to keep up with their studies.
In the United States, students are ranked relative to their classmates, not assessed against an external criterion, so they pressure each other not to study. Teachers are expected to pass almost all students, and if the class fails to study hard, the teacher is forced to lower the passing standard of the course. Subjects are taught at vastly different levels, but the rigor of the courses and the learning achievements that result are not well signaled to parents, neighbors, colleges, or employers, so rewards for setting difficult goals are small.
The French and Dutch models of secondary education combine in one system many of the most drastic reforms that have been proposed for the United States:
- Externally set subject-specific achievement exams taken by almost all secondary school graduates that supplement not displace teacher assessments of students. Grades on the external exams need to matter to the student, but they need not be the sole or primary determinant of desired outcomes, such as college admissions and access to the best jobs.
- Parental choice of upper secondary school and special field of study with money following students.
- Mastery learning with teeth. Those who fail two subjects in secondary school are required to either repeat the grade or transfer to a less demanding school or program.
- Secondary teaching is available only to those who demonstrate very high levels of competence in their subject. High entry standards are sustained by offering high wages and good working conditions.
- High standards for admission to the next stage of education.
This system of incentives and school organization appears to work for France and the Netherlands. A similar system, lacking only the externally set exit
exams, also works well in undergraduate education in the United States. At the secondary level, however, such reforms are controversial. Successful implementation of any one of these reforms would be a major political undertaking. Implementation of the whole package of reforms is not politically feasible at present. Yet the analysis here suggests that in Britain when just two elements of the package—mastery learning with teeth and attractive teacher salaries—were missing and a third element—school choice—was only recently introduced, achievement levels were substantially lower than in the Netherlands and France. Consequently, from a practical policy point of view, the message is not very positive. School climates and education standards do not change rapidly and easily. France and the Netherlands have not discovered a cheap and painless route to higher achievement.
The important lesson is that incentives, both their strength and structure, matter. There are less controversial ways of increasing the rewards for academic achievement, so the analysis here should not cause American reformers to despair. Reforms tailored to the American context have a greater chance of successful implementation than any effort to replicate the French or Dutch systems of secondary education.
President Clinton, former President Bush, and most of the nation's governors support the development of a system of European-style achievement examinations for upper secondary students. Everyone recognizes, however, that the decentralized character of American education and the controversial nature of specifying and assessing what young people should know and be able to do requires a slow, consensus-building approach. Consequently, it will probably be decades before external examinations in specific subjects are widespread in the United States. School cultures are resistant to change, so significant improvements in achievement will take even longer.
Lessons for Economic Analysis of Education Issues
Much of the economic research on elementary and secondary education has employed a production function paradigm. Conventionally, test scores measuring academic achievement are the outputs, teachers are the labor input, and students are goods in process. Even though I have written papers in this tradition myself, I am concerned that many of the inputs that conventionally appear on the right in these models are really endogenous and that severely biased findings may result.
This paper points in different directions. Schools are viewed as worker-managed organizations producing multiple products. In the classroom/school team production unit, students are as much workers as the teachers. Students are also consumers who choose which goals or outputs to focus on and how much effort to put into each goal. The behavior of each of the system's actors—teachers, administrators, school board, students, and parents—depends on the
incentives facing them. The incentives, in turn, depend on the cost and reliability of the signals that are generated about the various outputs of the system. The discussion above demonstrates the relevance of agency theory, game theory, signaling theory, and other elements of economic theory to the understanding of how schools and students operate, but it only scratches the surface. Deeper plowing of these furrows will yield a large crop of new insights into education and education policy.
References
Adler, M., and G. M. Raab. 1988. "Exit choice and loyalty: the impact of parental choice on admissions to secondary schools in Edinburgh and Dundee." Journal of Educational Policy 3:155–179.
Bishop, J. 1990. "Incentives to study: why American high school students compare so poorly to their counterparts overseas." Pp. 17–51 in Research in Labor Economics, vol. 11, D. Crawford and L. Bassi, eds. Greenwich, Conn.: JAI Press.
Bishop, J. 1993. "The Impact of Academic Competencies on Wages, Unemployment and Job Performance." Carnegie/Rochester Forum, B. Malkiel, ed.
Bradley, A. 1993. "Not making the grade: teacher firing spurs debate over standards and expectations for students." Education Week, Sept. 13, Pp. 1, 19–21.
Central Bureau Voor De Statistiek. 1993. Education Statistics. The Hague, Netherlands: Central Bureau Voor De Statistiek.
Cooper, H. M. 1989. Homework. White Plains, N.Y.: Longman.
Costrell, R. 1994a. "A simple model of educational standards." The American Economic Review 84(4):956–971.
Costrell, R. 1994b. "Centralized vs. decentralized educational standards under pooling." Department of Economics, University of Massachusetts, Amherst.
Ehrenberg, R., and D. Brewer. 1993. Did teacher's race and verbal ability matter in the 1960's? Coleman revisited. School of Industrial and Labor Relations, Cornell University, Ithaca, N.Y.
Elley, W. 1992. How in the World Do Students Read? The Hague, The Netherlands: International Association for the Evaluation of Educational Achievement.
Ferguson, R. 1990. Racial Patterns in How School and Teacher Quality Affect Achievement and Earnings. Kennedy School of Government, Harvard University, Cambridge, Mass.
Frederick, W. C. 1977. "The use of classroom time in high schools above or below the median reading score." Urban Education 11(4):459–464.
Frederick, W., H. Walberg, and S. Rasher. 1979. "Time, teacher comments, and achievement in urban high schools." Journal of Educational Research 73(2):63–65.
Gamoran, A., and M. Barends. 1987. "The effects of stratification in secondary schools: synthesis of survey and ethnographic research." Review of Education Research 57:415–435.
Goodlad, J. 1983. A Place Called School. New York: McGraw-Hill.
Government Statistical Service. 1993. Education Statistics for the United Kingdom: 1992. London: Her Majesty's Stationery Office.
Graham, A., and T. Husted. 1993. "Understanding state variation in SAT scores." Economics of Education 12(3):197–202.
Grissom, J. B., and L. A. Shepard. 1989. "Repeating and dropping out of school." Pp. 34–63 in Flunking Grades: Research and Policies on Retention, L. Shepard and M. L. Smith, eds. New York: Falmer Press.
Hanushek, E. A. 1971. "Teacher characteristics and gains in student achievement: estimation using micro-data." American Economic Review 61(2):280–288.
Hollenbeck, K., and B. Smith. 1984. The Influence of Applicants' Education and Skills on Employability Assessments by Employers. National Center for Research in Vocational Education, Ohio State University, Columbus.
Holmes, C. T. 1989. "Grade level retention effects: a meta-analysis of research studies." Pp. 16–33 in Flunking Grades, L. Shepard and M. L. Smith, eds. New York: The Falmer Press.
International Assessment of Educational Progress. 1992a. Learning Science. Princeton, N.J.: Educational Testing Service.
International Assessment of Educational Progress. 1992b. Learning Mathematics. Princeton, N.J.: Educational Testing Service.
Kang, S. 1985. "A formal model of school reward systems." In Incentives, Learning and Employability, J. Bishop, ed. Columbus, Ohio: National Center for Research in Vocational Education.
Kominski, R., and R. Sutterlin. 1992. What's Its Worth? Educational Background and Economic Status: Spring 1990. U.S. Bureau of the Census, Current Population Reports, P-70, No. 32.
Kulik, J., and C. L. Kulik. 1984. "Effects of accelerated instruction on students." Review of Educational Research 54(3):409–425.
Lewis, H. D. 1985. The French Education System. New York: St. Martin's Press.
Locke, E. 1968. "Toward a theory of task motivation and incentives." Organizational Behavior and Human Performance 3:157–189.
Longitudinal Survey of American Youth. 1989. Data File User's Manual. Dekalb: Public Opinion Laboratory.
Lundberg, I., and P. Linnakyla. 1993. Teaching Reading Around the World. The Hague: International Association for the Evaluation of Educational Achievement.
Madeus, G., and T. Kelleghan. 1991. Student examination systems in the European Community: lessons for the U. S. Report to the Office of Technology Assessment.
McKnight, C. C., et al. 1987. The Underachieving Curriculum: Assessing US School Mathematics from an International Perspective. Champaign, Ill.: Stipes Publishing Co.
Ministere de l'Education Nationale et de la Culture. 1992a. Reperes and References Statistiques sur les enseignements et la formation . Paris: Ministere de l'Education Nationale et de la Culture.
Ministere de l'Education Nationale et de la Culture. 1992b. L'Etat de L'Ecole. Paris: Ministere de l'Education Nationale et de la Culture.
Monk, D. 1992. Subject Area Preparation of Secondary Mathematics and Science Teachers and Student Achievement. Department of Education, Cornell University, Ithaca, N.Y.
National Center for Educational Statistics. 1992. The Digest of Education Statistics: 1992. Washington, D.C.: U.S. Department of Education.
National Center for Educational Statistics. 1993. The Condition of Education: 1993, vol. 1. Washington, D.C.: U.S. Department of Education.
National Center for Educational Statistics. 1994. Occupational and Educational Outcomes of Recent College Graduates 1 Year After Graduation: 1991. NCES 93–162. Washington, D.C.: U.S. Department of Education.
Nelson, H., and T. O'Brien. 1993. How U.S. Teachers Measure Up Internationally: A Comparative Study of Teacher Pay, Training, and Conditions of Service . Washington, D.C.: American Federation of Teachers.
Nijhof, W. J., and J. N. Streumer. 1988. "The Netherlands." In The Encyclopedia of Comparative Education and National Systems of Education . Oxford: Pergamon Press.
O'Leary, J. 1993. "Universities reduce A-level score needed for degree courses." The Times, July 12.
Organization of Economic Co-operation and Development. 1989. Employment Outlook. Paris: OECD.
Organization of Economic Co-operation and Development. 1992. Education at a Glance. Paris: OECD.
Organization of Economic Co-operation and Development. 1993. Education at a Glance. Paris: OECD.
Peter D. Hart Research Associates. 1994. Survey of Teachers Represented by the American Federation of Teachers.
Postlethwaite, T. N., and D. E. Wiley. 1992. Science Achievement in Twenty-three Countries. London: Pergamon Press.
Raffe, D. 1984. "School attainment and the labor market." Pp. 174–193 in Fourteen to Eighteen: The Changing Pattern of Schooling in Scotland , D. Raffe, ed. Aberdeen: Aberdeen University Press.
Robitaille, D., and R. Garden. 1989. The IAEEA Study of Mathematics II: Contexts and Outcomes of School Mathematics. New York: Pergamon Press.
Sizer, T. R. 1984. Horace's Compromise: The Dilemma of the American High School. Boston: Houghton Mifflin.
Stedry, A. C. 1960. Budget Control and Cost Behavior. Englewood Cliffs, N.J.: Prentice-Hall.
Strauss, R. P., and E. A. Sawyer. 1986. "Some new evidence on teacher and student competencies." Economics of Education Review 5(1):41–48.
Suskind, R. 1994. "Put down, kicked around, honor students struggle on." The Wall Street Journal, May 26.
Ward. 1994. "A day in the life." N.Y. Teacher. Jan.
Wiley, D. E. 1976. "Another hour, another day: quantity of schooling, a potent path for policy." In Schooling Achievement in American Society , W. H. Sewell, R. M. Hauser, and D. L. Featherman, eds. New York: Academic Press.
Willms, D. J., and F. Echols. 1993. "Alert and inert clients: the Scottish experience of parental choice of schools." Economics of Education Review 11(4):339–350.
Wood, R. E., A. Mento, and E. Locke. 1987. "Task complexity as a moderator of goal effects: a meta analysis." Journal of Applied Psychology 72(3):416–425.
This page in the original is blank. |