The Effects of School Choice on Curriculum and Atmosphere

Choice-based reforms are anchored on the idea that parents will have the power to make schools more responsive to them if they will have the ability to choose among these institutions since their choices determine enrolments and budgets. The main issue to this view centers around how and what kind of schools parents select. Little, however, is known on how they go about this course of action.

Aside from the method of choosing, school choice reforms also question school effectiveness, finance, student peer groups, and pupil achievement. Other questions are addressed to determine if choice schools offer high-quality education as suggested by test scores, college attendance, and other measurements. It is ambiguous, however, how choice schools yield higher achievement scores or greater school efficiency.  It remains to be revealed how schools react to the pressure that is imposed by parents.

Proponents and opponents of choice-based school reforms both use survey results to support their arguments by citing school attributes. Such evidence, however, is problematic because of the selection problem and the difficulty of eliciting truth from surveys with regards to preferences. The first one may offer misleading evidence for the prediction of the impacts of school choice policy since schools and parents may not be representative of the population. The latter comes from the complexity encountered in setting up realistic choices.

The answer to the question of how parents choose schools is challenged by four obstacles.

1. The selection problem.

2. The requirement for objective evidence.

3. The choice programs are determined with other policies.

4. The need to differentiate between the effects of student segregation and the effects of giving more influence to parents’ preferences.

The strategy of the author includes identifying disparities in the number of school districts by using topographic differences among metropolitan areas. This classifies exogenous differences in the degree of school choice among public school districts. Generous data on schools and pupils are used to obtain objective measures of behavior to verify subjective descriptions of school attributes. Data from the National Education Longitudinal Study (NELS) are used. Demographic data from the Public School Universe and the special school district tabulation of the 1990 census are also utilized to determine the impacts of segregation from the impacts of school choice.

 

I. Five Issues about Parents’ Choice Behavior

The argument as to how parents’ decisions affect schools is connected to five issues.

Proponents of choice claims	Opponents of choice claims
Parents are more likely to participate in decision making.	Parents are less willing to do so if they are able to pick another school.
Parents want their children to be challenged by their academic curriculum.	Parents prefer easy curriculum.
Parents like stringent academic environments.	Parents want environments that do not go against extra-curricular activities.
Parents like schools that impose strict discipline upon students.	Parents want schools with lax standards.
Parents prioritize academic preparation.	Parents give more weight to extracurricular activities.

II. The Interaction of Parents’ Preferences and School Choice

There are variations in the preferences among and between parents and teachers. The teaching methods of parents and teachers diverge most of the time because they have varying educational backgrounds. Teachers lean towards unpopular teaching styles with conservative parents. Also, parents favour teaching and disciplinary methods that are costly in terms of teacher effort, while teachers prefer methods that are costly on the basis of parent effort.

School choice impacts how the parent preferences are transmuted into school attributes within parent-teacher groups and between them; it will let more variation in preferences to be revealed. But, the predicted rise in variation of school traits is not the central concern of the study. Rather, the mean impact of school choice on curriculums and atmosphere in schools will be the focus of the discussion. If average parents like weaker standards than by the staff of the school, greater choice will lead to the fall of academic standards. Conversely, if average parents favour higher standards than by the staff of the school, greater choice will enhance academic standards.

III. Data

Extensive information on curricular and extracurricular activities is required for an empirical study on school attributes that parents prefer. Data must depict the decisions of individual schools and describe the environment where the school operates. Such precision, however, is not available on any source. As such, information about the student body and the community of each school district in the United States was derived from the National Education Longitudinal Study (NELS) and was matched with the Public School Universe and the special school district tabulation of the 1990 Census of Population and Housing.

The Public School Universe has administrative data on the pupils in each school and district. Educational attainment, household income, age, and house value were derived from this dataset.  The NELS, on the other hand, is comprised of data from 27,805 pupils from 2,451 schools in 1988. It consists of student, school, and transcript components, which present objective information and subjective assessments for the five parent choice behavior issues.

Most of the variables for curricular and extracurricular activities are from the school component of NELS. Meanwhile, parental involvement is assessed using five variables – the fraction of parents who discuss high school curricular choices with their children, who visit school, who attend meetings, who attend parent-teacher conferences, and the administrator’s assessment of how much parents are involved in decisions about school policy.

To assess whether a school encourages students to take a challenging curriculum, the proportion of students who took the advanced placement scores is taken, while the sternness of the academic environment is evaluated using six measures – an index of grade inflation in mathematics, an index of grade inflation in English, number of hours of homework assigned per week, number of standardized credits required for graduation, administrator’s assessments of the degree to which students place a priority on learning and are expected to do their homework.

The measurement for grade inflation in mathematics is obtained by taking the difference of the percentile score on a standardized mathematics test and the grade point average in mathematics. Grade point averages are measured on a four-point scale. The percentile scores are divided by 25 before subtracting so that they too will be on a four-point scale. Student differences are then averaged within the school. One letter grade’s worth of grade inflation is indicated by an increase of 1 in the school’s index. The index of grade inflation in English is computed similarly.

To determine if a school has a structured, disciplined environment, two objective measures are assessed – the number of minutes a pupil spends in class daily and the action taken by the school against a delinquent student – and four subjective measures the administrator’s assessment of the degree to which the school has structured classrooms, has a structured school day, does not tolerate deviation from rules, and provides a flexible environment.

The weight that a school puts on sports programs relative to academic programs is gauged using the administrator’s assessment of the degree to which the school emphasizes sports and the ratio of physical education to faculty in core subjects. The percentages of students who join varsity teams, intramural sports, band, choirs, and orchestras are also taken.

Descriptive statistics of the variables mentioned above reveal that parent involvement has a large variance. The percentages of students who take AP courses also vary extensively. And although about half of schools have a regular mathematics track that is challenging, great variation was also observed.

IV. Empirical Strategy

An intuitive strategy to identify the impact of choice on school policies would be to compare choice and nonchoice schools with students who are similar in terms of their measured aptitude and demographics by using linear regression on a school policy measure and on a variable that indicates whether the school is a choice school. However, this strategy is troubled by a serious selection problem. Parents and children who select choice schools are different from those who do not on the basis of both observable and unobservable traits. As such, parents and students that are expected to select choice schools might act differently even if they used regular schools and the policy in choice schools will reflect both policy decisions and different types of pupil.

Some researchers tried to eliminate the selection problem by evaluating choice at the district level rather than on the school level. A district is specified a choice district if it has one choice school. The evaluation between choice and nonchoice districts can show how policy is altered when schools are subject to more parental pressure because of choice. This type of comparison is normally done using linear regression of a school policy measure on variables that describe the demography of the school and a variable that indicates whether the district is a choice district. This strategy deals with the selection problem because a choice district must deal with all parents and students in the district, not just those who actively choose. However, one more problem remains: choice programs are usually reactions to underlying conditions in the district. Comparisons between choice and nonchoice districts cannot distinguish between the effects of having a choice program from the effects of the circumstances that caused the program.

To produce unbiased estimates of the effect of choice on school policy, a variation in the degree of choice that is independent of the circumstances of the school and district must be accounted for. Statewide programs are potential subjects for future research. However, it disregards variation due to differences in school districting – an important source of variation in the degree of school choice available in the United States.

Herfindahl indexes are a widely-accepted measure of the number and variety of school districts in a metropolitan area. This measure is constructed by first getting the school district’s share of total enrolment in the metropolitan area. Then these shares are squared, summed within each area, and multiplied by -1. The index will be equal to -1 if there is no choice at all and moves to zero as the number of school districts increases and as enrolment is spread evenly across districts. The lower the index, the less choice among districts there is in a metropolitan area. Also, the difference between metropolitan area with and without choice is approximately a one-unit change in the index.

The average metropolitan area in the United States is about -0.17. This measure, however, is highly variable. Also, parents are likely to concentrate in school districts that are good in unobservable ways. This biases results toward finding metropolitan areas with high enrolment rates. To reduce this bias, the researcher must focus on determining if a metropolitan area is likely to be divided into many districts and to ignore enrolment concentration that is caused by parents choosing more successful districts. This is done by estimating the regressions by instrument variables. Both Ordinary Least Squares (OLS) and instrumental variable (IV) estimates are presented.

To differentiate between the effects of choice and the effects of a homogenous student body, the measures of population homogeneity within districts are controlled in the regression. Data from the Public School Universe and the school district tabulation of the census are essential to this. They allow for the measurement of within-district homogeneity gauges.

V. Results

A. The Effect of Choice on Parental Involvement

Both OLS and IV estimates suggest that choice increases parental involvement. An increase in choice causes parents to visit the school more often, discuss the curriculum with their children, and trigger administrators to raise parental involvement in policymaking. Also, greater choice causes a higher percentage of parents to attend PTA meetings and parent-teacher conferences. These results imply that parents believe that they get more return from involvement when they have more choice. This is probably because as suggested by the bargaining theory, administrators are more willing to set policy that goes with parent preferences when the threat of exiting is more likely.

B. The Effect of Choice on the Curriculum

There is considerable evidence suggesting that, consistent with survey findings, schools that are under more pressure from parental choice expose their students to more intellectually challenging study. This infers that if given the choice, schools would rather give an easy curriculum because otherwise, school teachers and administrators will bear much of the work. This result is also consistent with a previous study, which discovered that students who come from schools that face more competition are likely to have higher achievement test scores.

C. The Effect of Choice on the Academic Environment

Schools that are in an environment where parents have more choice allow more grade inflation. This suggests that parents respond to incentives given by college admissions. This is under the assumption that admission officers encounter difficulty in adjusting for grade inflation. Or perhaps, employers and parents are misled by inflated grades. Note, however, that grade inflation does not mute incentives to perform. Results also show that schools are more likely to encourage students to develop good habits of work and learning when they are subject to greater choice. It appears that parents are capable of evaluating some school policies and of choosing schools that offer policies that are consistent with their preferences.

D. The Effect of Choice on a School’s Emphasis on Structure and Discipline

Consistent with the preference of parents, results indicate that Tiebout choice compels schools to emphasize discipline and give more structure. If left on their own, schools would rather implement less structure and discipline. This could be because the school staff will bear much of the cost associated with giving and maintaining more discipline.

E. The Effect of Choice on the Priority of Sports and Extracurricular Activities

There is no enough evidence to confirm that more choice for parents will expand the priority given to sports and extracurricular activities. Results do not indicate that choice suppresses the said activities. However, it also does not support the conjecture stating that choice would sacrifice academic programs to them.

VI. The Importance of Accounting for Selection and Endogenous Programs

It was shown that when selection and endogenous programs are not accounted for, results will be biased and difficult to interpret. They will be unreliable in indicating what average parents seek in a school. Comparison between choice and nonchoice schools and districts verified this point.

Comparing choice and nonchoice schools will be subjected to selection bias if parents and students who choose their schools differ in unobservable ways from the rest of the population.  Meanwhile, estimates that are generated from comparing choice and nonchoice schools will be biased if intradistrict choice programs are endogenous responses to district circumstances. As expected, all results show inconsistent patterns and are not usable for rational interpretation.

 

VII. Conclusions

Overall, results indicate that choice would not weaken academic and disciplinary standards in the U.S. Also, what parents say they want out of schools is what they choose given the opportunity to do so. Schools that operate in metropolitan areas where parents can choose easily among school districts tend to have more challenging curriculums, stricter academic requirements, and more structured and disciplined environments. Schools, on the other hand, do not focus much on sports and extracurricular activities. Choice seems to make parents more involved in what their local schools do.

On top of this, an interesting result was observed. It appears that greater parental choice cause grade inflation. This indicates that someone – college admission officers, employers, or parents – is fooled by nominal grades. Even so, despite lower nominal standards, the pressure of choice still causes schools to raise real standards. That is, students still achieve higher scores on standardized tests.

The goals of parents for schools are aligned with traditional ones. They want students to excel academically and have good work habits and self-discipline. With this consequence, school reforms that allow parents to challenge school policies like vouchers and charter schools, will most likely prompt higher academic standards.

Source:

Caroline M. Hoxby, “The Effects of School Choice on Curriculum and Atmosphere” In Susan Meyer and Paul Peterson (Eds.) Earning and Learning: How Schools Matter, Washington D.C.: Brookings Institution Press, 1999, pp. 281-315.