Showing posts with label Education. Show all posts
Showing posts with label Education. Show all posts

All School Finance Equalizations are not Created Equal

I. Introduction

School finance equalization (SFE) has affected American schools in the areas of school spending, private schooling, and achievement for over 30 years. Economists have become interested of SFE because optimal social investment in human capital depends on how school spending is distributed across students. However, in spite of its importance, SFEs are still poorly understood in practice.
State courts and legislators are confused, not on their goals or legal issues, but on how to implement such goals. States have relied on lawyers rather than economists for the implementation of programs. Owing to this, the schemes became manifestations of legal rhetoric of equalization rather than economic logic about taxation and redistribution.  As such, equalization debates and court cases recur and states have started to give up understanding school finance to identify trial-and-error methods of finding suitable schemes.
SFE schemes sometimes bring results that were not intended. They can reduce total school spending even if it is divided evenly (levelling down). Hence, greater equality in spending also brings lower average spending. In extreme cases, spending may possibly drop even in districts that were meant to be beneficiaries. Several commentators from the famous California SFE argued, however, that the spending reduction that they experienced was not a result of levelling down and was not associated with SFE.

II. School Finance Schemes

A. Pure Local Property Tax Finance
The core of most state systems remains to be local property tax finance. Districts spend their local property tax revenue and households divide their disposable income among property taxes, housing consumption and other consumption.

B. Categorical Aid
Before SFEs became popular, categorical aid was the most famous method of redistribution among    school districts. Unlike SFEs, categorical aid is financed by state income or sales taxes. It is also manifested on one or more district attributes – like mean household income or poverty rate – but not on property values as with the case of SFEs. This type of aid seeks to help school districts that are liquidity constrained or those that have children that are expensive to educate. With flat categorical aid, the disposable income of households is reduced by a school-related tax on income. Each district then receives a grant for each student that is a function of the income and demographics of households. Meanwhile, matching grant categorical aid schemes matches spending that were raised locally[1].

C. Foundation Aid
The most common type of SFE scheme is the foundation aid (FA). FAs are like flat grant categorical aids except that they redistribute not on socio-demographic characteristics of households but among districts based on per-pupil property values. FAs are usually self-funding, so that total foundation taxes that are paid equates to the total foundation grants disbursed. Moreover, FA programs are more stringent in states where the foundation grant is higher relative to per-pupil spending.

D. Power Equalization/Guaranteed Tax Revenue Schemes
Guaranteed Tax Revenue (GTR) and power equalization schemes are often used by states who intend to employ stringent equalization. These schemes attempt to make the same tax rate yield the same revenue for each school district in the state. Many GTR schemes provide equal revenue for each school district in the state. They also offer stronger redistribution for districts that have higher tax rates. These schemes are self-funding at times, where tax rates and guarantees are selected to make sure that contributions from high property value per pupil districts help those that have low property value per pupil. To prevent districts from bailing out of GTR schemes, states require a minimum local tax rate. GTR schemes make each district's tax price – the amount of revenue it has to raise in order to spend an extra dollar – a positive function of its per-pupil valuation. A district's tax price can be derived by differentiating its budget constraint with respect to its local tax rate.

III. School Spending Under Various School Finance Schemes

School finance schemes are not always mysterious. They can be characterized by a few variables that summarize their income and price effects. It was observed that they level down compared with categorical aid schemes that are supposed to achieve the same amount of redistribution. This is one of the consequences of having a scheme for redistribution among jurisdictions based on assets whose prices rest on how people regard the jurisdictions. If the few crucial variables are used, the conflicting evidence on the effects of SFE can be settled.
Both categorical aid and SFE can decrease the distribution of spending. Theory, however, does not indicate how categorical aid and SFE affect spending significantly. Whether categorical aid levels are up or down compared with local finance is still an empirical mystery. Even though GTR or matching grant categorical aid can influence tax prices, almost all actual variation in tax prices is generated by GTR schemes.

IV. Beyond the Direct Effects of Equalization on Spending

Beyond the direct effects of SFEs, it was observed that the consequences of SFEs can be capitalized. Also, some SFE schemes have integrated feedback loops that augment the chance of leveling up or down. Meanwhile, Silva and Sonstelie (1995) mentioned that an SFE may begin with mean state spending, but the target will eventually move towards the median level of preferred spending in the state because of the decisiveness of the median voter.
Households are having a hard time enrolling to their desired schools after an SFE may shift to private schooling. They might send their children to private schools or create local educational foundations that privatize parts of local schooling. SFEs may also affect student achievement by changing school resources and schoolmates.
It was detected that achievement is likely to increase more among low-performing students than among high-performing ones. Lastly, even an equalization that initially levels up can level down as it ages if voters respond to the SFE by constraining the state (through property tax limitations and targets that are strictly tied to inflation) so that future adjustments in the formula cannot further utilize their property wealth.

V. Data

Data on expenditure, enrolment, property values, property tax rates, socio-demographic characteristics, and information about each state's school finance laws were derived from the following sources:
  • Census of Government Data (1972, 1982, 1992)
  • Census of Population And Housing data summarized at the school district level
  • Consolidated districts and special districts
  • Property values from statistical abstracts as well as data from the Census of Population and Housing
  • The Public School Finance Programs series I derived basic information on school finance laws from the
  • Series of Public School Finance Programs (1968-1969, 1970-1971, 1978-1979, and 1987-1988 school years)
 VI. Facts about School Finance Equalizations

Close examination of SFEs from 1970 to 1990 reveals interesting facts. First, the incentives imposed were not accurately revealed by the categorization into court-ordered, legislative, and "no equalization". It was also observed that the majority of states in the "no equalization" category have equalization activity. This activity suggests that judicial or political situations can make an SFE known, even if it is more modest than other states' routine changes to their school finance laws. Furthermore, some court-ordered and legislative "equalizations" change parameters in an anti-equalization direction.
Such behavior makes one doubt whether a dummy variable for equalization will correctly represent the economic content of school finance schemes or the usefulness of an empirical strategy with two dummy variables.

VII. Estimation Strategy

The values of the school finance variables were calculated for every district in the United States in 1970, 1980, and 1990. This brought all of the schemes into a common framework. Regression analysis was utilized to estimate the effect of SFE on the following outcomes:
· The level of per-pupil spending
· Inequality in per-pupil spending
· Property values
· Private school attendance
· Student achievement
An education production function was estimated in the last case using the high school drop-out rate as the measure of student achievement and SFE as a shock to districts' spending. Endogeneity of the observed values of the school finance variables was accounted by simulating the values each district would face if it were to have remained at its pre-SFE tax rates, spending, and property values.

VIII. Spending Results

Two important findings were detected:
1. Changing the inverted tax price that a district encounters can cause dramatic changes in spending (the magnitude of the effect of the inverted tax price); and
2. Flat grant categorical aid has almost no potential to generate levelling-down
After calculating what mean per-pupil spending would have been if no equalization scheme were in effect, it was observed that equalization schemes have generated both levelling-up and levelling-down. Meanwhile, results for the pupil-weighted coefficient of variation in per-pupil spending for each state in 1990 reveal that the standard deviation of New Hampshire's per-pupil spending would have been 5.6 percent more of its mean per-pupil spending if there had been no equalization program. Furthermore, the log 90-10 ratio[2] shows that students at the ninetieth and tenth percentiles would face spending that differed by about 59.3 percent if there had been no equalization scheme.
New Hampshire had a moderate FA scheme in 1990. Majority of states have equalization experiences as moderate as that of New Hampshire. However, although the most common equalization experience in the United States is moderate, it is not necessarily optimal. Among states that experience similar equalizing, there is still variation in whether the state is levelling down or up.
The states with the most dramatic equalizing are all levelling-down[3]. In these states, the difference between spending at the ninetieth and tenth percentiles would be almost 20 percent greater if no equalization program were in effect. In the states with the greatest increases in mean per-pupil spending[4], less dramatic equalizing was detected. In these states, the difference between spending at the ninetieth and tenth percentiles would be only about 10 percent greater if no equalization program were in effect.
Levelling-up schemes tend to have more simple equalizing than levelling-down schemes because these states face little or no direct cost when they set inverted tax prices close to zero but face substantial costs when they set inverted tax prices higher. It is costly to bribe districts that would prefer low spending into spending a lot but it is inexpensive to forbid high spending.
Usually, courts and legislatures are required to achieve a combination of equity and adequacy in school finance. To detect whether a scheme is achieving a good mix of equity and adequacy is the criterion: "Do no harm to poor districts." This means that if in pursuit of equity goals, a scheme makes mean spending fall significantly that spending is less sufficient in poor districts, it has almost certainly gone too far. 

IX. Other Effects of School Finance Equalization

SFE schemes can change property values, both because SFE schemes tax school productivity that has been capitalized in house prices and because SFE schemes themselves can be capitalized. Meanwhile, the most extreme leveling down schemes[5] raise the share of students who attend private school by about 3 percentage points. Furthermore, the private school share falls by about 1.5 percentage points in districts that have the inverted tax prices of 1.5. In general, the private school share will rise in "property-rich" districts and fall in "property-poor" ones.
For dropout rates, it was found that that dropping-out is affected only in districts that would otherwise be very low-spending. This will only occur in districts that were forced to boost spending by the imposition of the flat grant. Results imply that equalization enhances student performance the most in schools that would have very low spending if left to their own.

X. Conclusion

The findings show that some spending equality can be obtained with leveling-up or mild leveling-down. However, near equality of per-pupil spending cannot be achieved without very strong incentives. In theory, either leveling-down or leveling- up SFEs could incorporate such strong incentives, but, in practice, they are included only in leveling-down SFEs because it is cheaper to prohibit high spending but it is expensive to bribe districts to spend more than they prefer to.
Surprisingly, there are poor districts in the United States that might have higher spending if their state attempted to achieve less complete equality. Because equalization schemes can contain feedback where they reduce their own spending targets, underprivileged students can be worse off under equalization schemes that are more generous than the categorical aid schemes they replaced.
House prices drop in districts that are reprimanded for having high property value per pupil. The converse is also true. Such capitalization somewhat undoes the impact of a school finance equalization scheme. Parents are persuaded to send their children to private schools in districts that are penalized for having high property value per pupil.
Even though the estimated effects of equalization on student achievement are weak, it appears that the drop-out rate falls in districts that are constrained to increase spending by the imposition of a per-pupil spending floor.
These results imply that states should be careful in choosing their equalization schemes in order to move toward their equality and adequacy goals. The outcomes also hint that school districts react like economic agents to the incentives created by equalization schemes. Additionally, school finance equalization schemes generate unintended consequences because they centre redistribution among jurisdictions on assets (property) whose prices rely on how households value the jurisdictions. School finance equalization does not merely redistribute from people with greater to people with lower ability-to-pay. Another consequence is that it redistributes from schools that are more productive to schools that are less productive. It can also redistribute from people with greater taste for education to people with lower taste for education. Over time, its own formulation gets undone and a feedback that amplifies its intended effect is generated. Categorical aid, on the other hand, does not produce such unintended consequences.



Source:
Caroline M. Hoxby, “All School Finance Equalizations Are Not Created Equal” Quarterly Journal of Economics, Vol. 116, No. 4 (November, 2001), pp. 1189-1231.


[1] The matching rate is a function of district demographics
[2] The log of the ratio of per-pupil spending at the pupil-weighted ninetieth percentile and per-pupil spending at the pupil-weighted tenth percentile
[3] California, New Mexico, South Dakota, and Utah
[4] New Jersey, New York, Pennsylvania
[5] Schemes that lower the inverted tax price from one to zero

The Effects of Class Size on Student Achievement: New Evidence from Population Variation

I. Introduction 

Class size reduction among schools received attention – and budget – in the United States. It became popular as parents favored small classes because it enables teachers to use the teaching method they prefer. Teachers also find them favorable since they call for less effort in delivering instruction. For teachers’ unions, small classes are beneficial because they augment the demand for teachers. Meanwhile, administrators find reduced class size well-disposed because they expand the size of their domain.
Class size, however, is an example of the education production function erroneous belief, which suggests that educational inputs systemically transform into achievement. This analogy does not hold for the reason that the firms’ production functions are not merely a result of their ability to turn inputs into outputs; it is a consequence of maximizing an objective given a production possibilities set. Class size is also difficult to analyze since variation in class size stems from the various choices that are made by parents, schooling providers, or courts and legislatures. Class size is correlated to other variables of student performance making the results biased; analysts, however, hardly ever explain where the variation really comes from.
Experiments are rarely explicit. The majority of them take place in developing countries and most of the time, primary participants are aware that there is an undergoing experiment. Some players of the study are cognizant that if the experiment becomes unsuccessful, the education policy will not be passed. Hence, some schools append incentives that will not be included when the policy is actually passed. Some individuals also enhance their productivity when they are being evaluated (Hawthorne effect). In addition, a few experiment participants attempt to undo the randomness of the experiment by assigning the best students or teachers in the test classes.
These empirical and economic concerns are addressed in this study using two identification strategies, both of which use variation in class size that comes from population variation. The first strategy uses natural randomness in the population and straightforward logic – making use of variation from the random timing of births, which corresponds directly into differences in class size between cohorts. The said identification strategy is employed by separating the random component of population variation using long panels of data on enrollment and kindergarten cohorts in Connecticut school districts. The long panels are capable of eliminating all smooth changes in population. Residuals that remain after fitting a quartic function of time are then used separately for each grade in each school.
The second strategy makes use of the fact that class size abruptly shoots when a class has to be added to or subtracted from a grade when enrollment triggers a maximum or minimum class size rule. This identification strategy acknowledges that there is a discontinuous connection between class size and enrollment at certain known levels of enrollment while there is a smooth relationship between achievement and the determinants of enrollment. The panels of data are utilized to detect small changes in enrollment that are associated with changes in the number of classes in each grade in each school. Meanwhile, each district’s class size rules are used to ascertain if adjustments in the number of classes are just the result of small changes in enrollment triggering a rule. This strategy is accomplished through comparing the class size and achievement of adjacent cohorts that go before and after each event.
When using population variation, the range of class size is conveniently the range that is relevant for policy. Also, schools are examined under the normal conditions that they usually experience.

II. Sources of Variation in School Inputs and the Potential for Bias

The major source of variation stems from school choice as a result of residence selection. Between-district school input differences that result from parent choices yield upward biased estimates of the input efficiency. Systematic variation within a district over time faces a similar effect. Simple comparisons of schools using cross-section data or time-series data may likely generate bias in favor of class size reductions since such reductions will seem to be more efficient than they really are if parents who are more involved to their children’s learning also choose school districts that offer smaller class sizes.
Empirical results suffer from different biases depending on the source of variation of the employed school inputs. Bias is ambiguous because parent involvement can either be compensatory or reinforcing. The bias generated by administrators and teachers can also be vague because it depends on whether they lean towards meeting parents’ demand for higher learning or attending to children with learning problems. Inputs are also determined by state judges and legislators who have the tendency to favor compensatory policies than reinforcing policies.
However, this paper attempts to explicitly formulate a source of exogenous variation, not just eliminate sources of bias.

III. Empirical Strategy

A general education function uses the natural log of class size, a vector of cohort indicator variables, a vector of school indicator variables, and a vector of observed student, parent, and community attributes to determine student achievement. Test scores, the common measure for student achievement, are divided by the standard deviation of students’ scores on the test to help understand unfamiliar test scores and make comparisons across different studies. Meanwhile, the natural log of class size is used to incorporate the fact that a one-student reduction is proportionately larger from a base of students. The vector of cohort indicator variables is taken into account to allow for tests that vary annually and to allow for teachers who adjust their teaching to a test’s content. The vector of school indicator variables is added to control for schools populations’ attributes that are constant across time. Lastly, the vector of observed student, parent, and community characteristics normally includes variables that describe variables such as the racial composition, educational attainment, and income of students and households.

1. The First Identification Method
Actual enrollment has a deterministic component – enrollment if the timing and number of births were a deterministic function; and a random component – variation in enrollment resulting from the fact that biology causes random variation in the timing and number of births. It is expected that the random component affects regular enrollment proportionally. The first identification strategy is obtained by doing the following:
a. Get estimates of the random part of enrollment variation;
b. Use the random variation in enrollment to identify random variation in class size;
c. See how achievement is affected by random variation in class size.
This method makes use of the fact that the observable student, parent, and community characteristics and the random component change much more uninterruptedly than enrollment in a specific grade-school-time does. However, since parents can respond to the class size of their child, the method allows for little bias. They can decide to transfer their children to another school, making the observable characteristics endogenous to the random component.
At the district level, transfers among schools within the district will cancel out. This will not completely eliminate the possibility of bias due to the fact that parents can still take their children and transfer them. The said bias, fortunately, can be removed by using the potential kindergarten cohort at the district level (‘‘K5’’) as the source of random variation in class size. They will, however, be stronger instruments for class size in early elementary grades than in later elementary grades because student mobility weakens the correlation between kindergarten cohort size and later grades’ cohort sizes.
The costs and benefits of adding another class rely not simply on how much local parents care about schooling but also on actual enrollment. If the procedure for the first identification method is carried out and the changes in the number of classes discounted, an increase in enrollment will reduce class size if it activates an increase in the number of classes. This can be remedied by having an indicator variable for each combination of a school and expected number of classes instead of having school indicator variables in the first- and second-stage equations.

2. The Second Identification Method
Unlike the first identification method, the second identification method takes advantages on changes in the number of classes and the fact that changes in the number of classes in a grade can produce sudden changes in class size. The simplest way to use these discontinuities is the cross-section method of exploiting maximum class size thresholds.
The second method is independent of the identification that comes from using the log of residuals as an instrument for class size; this makes the two methods as checks for each other. Also, predicted class size is also not a valid instrument except when the rule causes a change in the number of classes; estimates will be valid if identification depends only in discontinuities. Finally, identification develops only when the rule binds, so if one uses a rule that binds only in some schools, the effects of class size will only be applicable for those schools.
Because there are issues about identification based on discontinuities, changes in the number of classes that are created by small within-school changes in enrollment that cause a district’s maximum or minimum class size rule are used. The said method is more accurate and less prone to bias than the cross-section method because one can follow enrollment in a grade in a school over time and observe small changes in enrollment. It also compares adjacent cohorts within a school and uses variation from different districts.
The second identification method involves the following procedure:
a. Identify all of the events in which a school increased or decreased the number of classes in one of its grades;
b. Given just the change in enrollment and the district’s maximum and minimum class size rules, identify all the events in which the change in the number of classes was predictable;
c. Within the subset, maintain the events in which the change in enrollment that induced the change in class size was smaller than 20 percent;
d. Estimate a first-differenced version of the achievement equation using just the cohorts immediately before and after each event.

IV. Data

Data on elementary grades are the ones looked-for owing to the fact that the integer nature of teachers and classrooms is useful to relate natural population variation into variation in class size and composition. Elementary classes are less divisible than secondary school classes since elementary school instruction usually involves one teacher spending hours of each school day with a regular group of students in one classroom. Furthermore, class size is also necessary in elementary schools but poorly defined in middle and high schools, where pupils attend different class sizes in different subjects. Moreover, elementary schools are not large in comparison to higher school levels. Lastly, since school cohorts are defined by birth date, data on population-by-age at the school entry cut-off date is needed.
The Connecticut school data that was explored consists of 649 elementary schools from 146 elementary districts. Overall, 25 percent of the schools have typical cohort sizes smaller than 46 students; 50 percent have smaller than 63 students; and 75 percent have smaller than 92 students. Since 1986, Connecticut has administered state-wide tests in the fourth, sixth, and eighth grades annually. All data are available publicly and were obtained from the Connecticut Department of Education or its publications.

V. Some Illustrations

Graphs of the graduating class versus enrollment of schools that have different classrooms per grade show the variation that is useful for the first identification method. Meanwhile, the graphs of two of the three schools illustrate the variation that is useful for the second identification method. However, these graphs cannot be used officially to confirm whether there is a significant relationship between achievement and class size. Regression analysis must still be conducted.

VI. Results

1. Results from Commonly Used Methods of Identification
Conventional methods are inclined to yield estimates that are biased by correlation between class size and unobserved parent and community attributes. Parents with unobserved good characteristics are more likely to select schools with small class sizes and communities with unobserved good characteristics. If the estimates are to be interpreted, it can be concluded that the a 10 percent reduction in class size in grades 1 through 3 improves fourth grade math scores by 0.1468 (about 15 percent) of a standard deviation and a 10 percent reduction in class size in grades 1 through 5 appears to improve sixth grade math scores by about 13 percent of a standard deviation.
Examining district-level variables controls the observed parent and community characteristics. They reduce the estimated effect of class size on test scores, but the estimates are still all negative in sign, and two of the six estimates are statistically significant at the 10 percent level. Results are of mixed or minimal statistical significance, and the impacts are small. Observed demographics are controlled but it is uncertain whether the variation stems from exogenous sources.
Value-added specifications are thought to control for all the effects of family background and neighborhood. Such claims are invalid as value-added estimates may be biased either negatively or positively. However, it is likely that the preponderance of the bias favors class size appearing to be effective.
Conventional methods try to eliminate one source of suspect variation at the expense of having more ambiguous sources. When considering a policy variable like class size, it is better to start with sources of variation that are known to be exogenous and work from there.

2. Results from the First Identification Method
Results suggest that a random rise in enrollment increases class size. It suggests that enrollment residuals are strong instruments for class size. However, it does not affirm that smaller class sizes produce achievement gains. The estimates are mixed in sign, and none is statistically significant at the 5 percent level. Moreover, the estimates do not accept the hypothesis that class size reductions are more efficacious in districts that contain low income or African-American students.

3. Results from the Cross-Section Regression Discontinuity Method
The initial result shows that decline in class size improve achievement significantly. More specifically, a 10 percent reduction in third grade class size augments fourth grade math scores by about 12 percent of a standard deviation. As discontinuities are narrowed in, however, the variables become insignificant and some reverse their sign. Therefore, the initial statistically significant results are generated not by the discontinuities in the predicted class size function, but by the suspect parts of the function.

4. Results from the Second Identification Method: (The Within-School Regression Discontinuity Method)
The within-school regression discontinuity method is quite powerful despite the fact that it depends purely on discontinuous changes in class size driven by changes in the number of classes because it compares adjacent cohorts in the same school, who have little reason to be different apart from their different class size experiences. It also produces standard errors so small that if a 10 percent reduction in class size were to change test scores by just 2 to 4 percent of a standard deviation, the change would be statistically significant at the 5 percent level. Despite the small standard errors, none of the estimates is statistically significantly different from zero at the 5 percent level. This suggests that the estimated effects of class size reductions are rather precisely estimated zeros.

VII. Interpretation

The two identification methods show that class size reductions have little or no impact on achievement. They are independent and they serve as checks for each other. Their results are also robust to specification changes.
The findings are based on class sizes that range to between ten to thirty students per class. This is the appropriate range for American schools. To extend the results to classes with more than thirty students – the usual class size of most developing countries – might yield incorrect interpretation. Similarly, it will also be a mistake to extrapolate these results to classes with less than ten students.
In contrast to the results of this paper, Kruger (1999) observed that a ten percent reduction in class size for three years improves scores by about 13 percent of a standard deviation and a 10 percent reduction in class size for five years improves scores by about 17.5 percent of standard deviation. The difference results may be due to the fact that the natural experiment changed class size but did not altered incentives, while the policy experiment changed class size and contained implicit incentives for teachers and administrators to make good use of smaller class sizes. If this assumption is correct, then class size reduction policies should contain built-in evaluation and incentives.
Since Connecticut school personnel were unaware of the natural experiment, they did not have the chance to react to or alter the evaluation. Other policy experiments may be different because of Hawthorne effects or other reactive participant behavior.
Some of the difference in results might be attributed to the fact that teachers experience small class sizes repeatedly, but not every year. They may not modify their primary classroom style significantly when they have the opportunities presented by a smaller class. Class size reduction should be combined with instruction for teachers that helps them modify their teaching techniques. However, even if she does not lecture differently to a smaller class, a teacher can devote more effort to each student during every teaching activity that has an individual element.

VIII. Conclusions

This study uses natural variation to identify the effects of class size on student achievement. This method gives three benefits:
a. The variation in class size is realistically exogenous;
b. The participants in the natural experiment were not aware of being evaluated or mindful of rewards being contingent upon the outcome; and
c. Natural population variation generates fluctuations in class size that are in the range relevant to current policy.
The study used two methods, which both show how population variation can be used to consistently estimate the effect of class size on student achievement. The first method is based on segregating the random component of the natural variation in population for a grade in a school. Meanwhile, the second method is based on using the discontinuous changes in class size that occur when a small change in enrollment induces a maximum or minimum class size rule. Both methods produce results that are appropriate for considering class size changes in the range of 10 to 30 students.
Under the said methods, it was found that reductions in class size have no effect on student achievement. The estimates are so precise that a 10 percent reduction in class size improved achievement by just 2 to 4 percent of a standard deviation. Moreover, there is no evidence that class size reductions are more efficient in schools that contain high concentrations of low income students or African-American students.
The results do not suffer from bias due to omitted variables and endogeneity than are typical estimates that depend on variation in class size that is generated by parents, teachers, administrators, or policy-makers’ decisions. Also, unlike other studies, participants included in the study are not aware of being evaluated. Incentives did not alter the results of the experiment. The experiments represent actual class size reduction policies, which hardly include evaluations or incentives for schools to make good use of the opportunities provided by smaller class sizes.



Source:
Caroline M. Hoxby, “The Effects of Class Size on Student Achievement: New Evidence from Population Variation” Quarterly Journal of Economics, Vol. 115, No. 4 (November 2000), pp. 1239-1285.

Benevolent Colluders? The Effects of Anti-trust Action on College Financial Aid and Tuition

I. Introduction

From 1989 to 1991 the Departments of Justice (DOJ) looked into several private, exclusive colleges for price fixing. The investigation eventually concentrated on the Overlap group of 23 colleges which is about half of the most exclusive colleges in the United States.[1] Even though the Overlap group denied the price-fixing accusation, in 1991 they discontinued their regular spring meetings which dealt with calculations of pupils’ neediness, and stopped gathering ever since. The eradication of the Overlap meeting offers a rare chance to test if the need-based support is mainly a charitable movement or a way to internalize externalities among pupils, in which colleges produce students’ desired environments rather than the conditions pupils would actually undergo if they acted individualistically. A lot of students would like his college to keep a policy of providing aid grounded on the need for others while excluding him from the policy, but by having this, the college’s student body and ability to attract donations are affected.
One of the goals, therefore, is determined if the same pupil would prefer a need-based aid regime, knowing there are no exemptions or a regime wherein all pupils get the exemptions he desires for himself.

II. An Economist’s View of the Antitrust Case – Part One: A Review of Theory

The antitrust case against the Overlap colleges stressed the economic logic of both the district and appeal courts.

A. The Market for College Education
For economists, students act as both the consumer (by using up the good and services) and producers (by investing in human capital) of college education. Such role is hard to distinguish.
To some extent colleges differ in their technology for producing education. Extremely selective colleges have technology wherein students must have comparable preparation and skill so that they can interact and facilitate each other's education. On the other hand, less selective colleges use a technology wherein pupil dealings are not that vital, still well-prepared and competent students are important since they act as role models and raise the college's attractiveness to draw in good faculty.
Having students acting merely as consumers of education, colleges would plainly provide tuition, goods and services bundles, and pupils would just select among those packages until the market for college consumption cleared. Likewise, if students are employed by colleges as inputs into the production of education, then colleges would just provide salaries and pupils would, choose among them. Since pupils must consume and produce at the same college, colleges' packages, and salaries should concurrently clear the markets for pupils serving as consumers and as producers.
The students are expected to produce offers of comparable value at various colleges, albeit the packages may differ in terms of composition. On one hand, due to education production technology that works best when students are mixed with peers with equal producing value, several colleges end up having pupils who are somewhat homogeneous in their values as producers, and as a result, such colleges supply their entire students with the same bundles. On the other hand, colleges using other production technologies wind up with pupils who are heterogeneous in their value as producers and as a result, provide very heterogeneous salaries. Since their valued pupils will be given roughly the same goods and services as their other students (who contribute less as producers), the “wages” of quality pupils must vary from those of their peers so to clear the market.
Meanwhile, for the following discussion, students are classified by their value as producers and not by their characteristics that may augment their value.

B. Merit Aid
Up to now the study has been looking at the notional tuition and notional wage the college would be willing to enroll a student as a consumer and be willing to pay to each pupil as a producer, respectively. Colleges that are highly selective bear high notional tuition but also have students serving as important producers who have high market wages. Because these colleges have to stringently allocate admission, their list tuition (the amount that college call “tuition” will be referred to as “list tuition) is conspicuously way below the notional tuition that would clear the market for likely consumers of their education. Their list tuition nets out a major portion of the wage they offer their students. The colleges are able to impose list in equilibrium by taking in only pupils having high market wages.
In order to maintain its market equilibrium, merit scholarships are the discounts that individual students must have from each college’s list tuition.  A series of list tuitions and merit scholarships are market clearing if their net is equivalent to the differences of notional tuitions and wages.[2]
However, there is greater possibility for notional wages to appear as merit scholarships at colleges that do not have admission to guarantee that their pupils are comparatively homogeneous. At relatively non-selective colleges, notional and list tuition are the same and important pupils can be encouraged to enroll only through high merit scholarships. At selective colleges with somewhat homogeneous students, setting tuition for those who did not contribute to the production process and then giving every enrollee an equivalent merit scholarship would be pointless. Lastly, highly selective colleges charging list tuition at a level that contains their usual student’s wage have the difference of notional tuition and notional wages for their student to be often negative.

C. Banning Merit Scholarship
On account of merit scholarships, valuable pupils are drawn in less selective colleges, where they’d atypical compared to the other students. A common principle is that merit scholarships are crucial in keeping equilibrium among heterogeneous students within colleges, while list tuition and college admissions procedures preserve equilibrium between colleges. The disparities among student’s merit scholarships have a significant role in letting nonselective colleges to effectively attract highly valuable pupils; in contrast, it has little effect on highly selective colleges.
There would be significant effects if a college having non-selective admissions opted to use merit scholarship to manage a heterogeneous student population. On the other hand, little change would occur, if a college having highly selective admissions offered wide ranging tuition subsidies and provide small merit scholarship to deal with the small variations among its pupils.

D. Needy Students
Given a perfect capital markets for investing in human capital, underprivileged students would make optimal investments. Meanwhile, imperfect capital markets may induce disadvantaged pupils to make suboptimal investments in their educations while cause society to experience slower growth  owing to their decisions impact on other students’ human capital acquisition. Pupils who are under investing are unlikely to be optimized as producers and they would also provide a college too small incentive to supply education quality, and too much incentive to decrease cost. Thus the aid for underprivileged pupils can raise and encourage efficient human capital investments.

III. An Economist’s View of the Antitrust Case – Part Two: Financial Aid and the Overlap Procedures

A. Financial Aid
When a student applies for financial aid, a major measure is his Expected Family Contribution (EFC) which attempts to identify how much a family allot for a student’s college education on the basis of income, assets, other family members’ college expenses, and other expenditures like health care costs. EFC together with the expenses of attending a particular college the student goes through are used by the federal government in regulating federal scholarships such as Pell grants and the amount of federally guaranteed loans for which a student qualifies.
EFCs show what the students need but they do not decide what intuitional scholarship the student would be getting. Also, a college may decide to provide scholarship to a pupil, but the amount would be at the college’s discretion. Many colleges consider a student’s EFCs (the willingness of the family to pay their EFCs) in their admission decisions so that they are not bombarded with scholarship requests for admitted students who can’t afford the costs.
Worthy of notice is the federal government’s promotion of the use of need-based policies by mandating that any student receiving one dollar of federal support must only get need-based support. As a result, colleges that make extensive use of merit aid regardless having needy student population will be taxed.
Colleges in the Overlap group and some outside the group willingly committed themselves in offering each admitted student a guarantee that he would be able to afford the cost of attending if his family settled it’s EFC, he was eager to work for a modest salary, and he was willing to take out modest loans. Colleges that made this pledge supplied an aid package to every pupil that settled the entire gap between the cost of attending and his EFC.  Finally, these colleges committed themselves to provide grants based on the student’s needs – more grants to those who are gravely needy since they have lower capacity for work compared to less needy students.

B. The Ostensible Purpose of the Overlap Group
Note that such commitments caused the students’ needs to be the colleges’ financial obligations.
Some of differences in the Overlap group’s calculation of EFC with that of the federal government can be due to the fact those colleges’ voluntary obligations generated rules which could be worked around by financially convoluted parents. Students enrolled in selective private colleges are suspiciously from families having a high degree of financial sophistication wherein a significant fraction of their income is gained from capital. In contrast to wage and salary earnings, capital income is flexible and easy to conceal for a short period such as the time when children are in college. Therefore, for the calculation, Overlap colleges decided not to automatically allow capital income losses, contrasting that of the federal calculation which enabled it due to federal tax purposes. Basically, these colleges are against the increase of their financial obligation owing to parent who uses tax shelters to mask their income.
The last difference between Overlap and federal calculations was that the former’s formula “respected” other colleges’ charge to older siblings.[3]  This reduces the risk exposure of colleges, which would have been otherwise open to extensive variation in their obligation to each student.
Meetings held by colleges in the Overlap group allowed the sharing of ideas regarding what established modest amount of work and loans, given that aid bundles did vary in composition among Overlap colleges. Also, students could and did use greater aid packages as a strategic advantage to acquire more grants from competing colleges.
Although the Overlap meeting offered no enforcement mechanism to restrain any deviant colleges, it would make it hard for colleges to pursue one policy while claiming to pursue another without the deception becoming noticeable.
Colleges are concerned that without the Overlap calculation and meeting, their pledge to fully accommodate the needs would be taken advantage by cleverer parents who are less poor, and their judgment over how to interpret confusing parental data would in the end make the formula of need too subjective that a college could alter its policy while claiming not to have done so, thereby easily allowing free-riding. They proposed that if need-determined aid became less feasible due to the subterfuges, they may probably a call for the abandonment of their full coverage of need, their policies of taking in individual without consideration of need, and their varied student populations.

IV. An Economist’s View of the Antitrust Case – Part Three: DOJ’s Allegation

DOJ’s allegation was that the Overlap meeting enabled involved colleges to increase list tuition and to gain higher tuition revenue.
The first of the many reasons behind DOJ’s allegation was the fact that DOJ relied on theory for firms in which consumers do not produce and employees do not consume, without modifying the theory in terms of higher education. Second, it relied on standard arguments regarding price fixing and price discrimination without making adjustments for higher education. This caused practical problems for DOJ. Under standard assumptions, arrangements to let people’s fees to rely on their need would form price fixing and price discrimination. As a result, federal polices would be the key anticompetitive force in higher education. The federal government would act as the main organizer of need calculations, and would punish colleges that disregard a student’s need and in lieu grant merit aid, and fixes the tuition paid by tens of thousands of Pell Grant receivers. Federal policy is anchored on the notion that underprivileged students are extremely price-sensitive when they invest in education, and that decreasing such sensitivity would rectify their cost-benefit computations, wherein these adjustments rationalize any probable distortions to their human capital investments. The twofold argument is that underprivileged students’ extreme price sensitivity will resolve unwarranted cost- and quality- decreasing incentives that their conduct would otherwise provide colleges, and that these rectifications explain any plausible distortions to colleges’ cost and quality choices. If DOJ were to admit the validity of these arguments, then it would also have to admit that need-based support does not directly amount to price fixing and price discrimination. On the other hand, if it were to be refuted by the DOJ, then it must denounce the Overlap group. The DOJ’s attempts to get away from such binds are through giving support to need-based aid while claiming that the Overlap meetings were pointless for administrating need-based aid; and making only ephemeral appeals to conventional wisdom with regards to price fixing and price discrimination. 
Lastly, it is difficult to restructure DOJ’s logic since DOJ argued that the Overlap meeting established per se price fixing, for instance there are two firms regularly meeting and agreeing on prices and in such cases, DOJ is not obligated to prove that the firms use certain methods to increase prices or to expose any impacts on price.

A. The Allegation that the Overlap Procedures Helped the Colleges Collude on List Tuition
The most direct allegation by the DOJ was that the Overlap meeting aided in the colleges’ collusion on list tuition. If DOJ was right then the meeting’s effect is expected to be seen on the level of its list tuition, and that tuition announcements are made after Overlap meeting.  In addition, if non-Overlap colleges did not actively conspire but did seek for price leadership from the Overlap group, then it is expected to witness tuition announcements from non-Overlap colleges after the meeting.

B. The Allegation that the Overlap Meeting Allowed Colleges to Obtain Excessive Tuition Revenue
The primary allegation was that the Overlap meeting let the colleges gain more tuition revenue. DOJ suspected that the colleges were taking in students too cheaply – this implies that Overlap meeting permitted the colleges to offer students too little need-based policy and too little merit aid.  Therefore, removal of meeting would cause colleges’ admissions and aid polices to be more costly.
Moreover, the DOJ also argued that the eradication of the meeting would not alter the colleges’ student population’s neediness. But this would logically obliged the colleges to have rent, so to put an end to their meeting, maintain their policies, and just impose lower fees to students. It was also assumed that the colleges would choose their aid policies with no sensitivity to its price.

C. The Allegation that the Overlap Meeting Gave too Little Need-Based Aid to Needy Students
DOJ suspected that the Overlap calculation as well as the meeting that implemented it permitted the colleges to offer lesser need-based aids to underprivileged individuals.
A large portion of the DOJ case was devoted to proving that the Overlap formula provided pupils higher EFCs than their federal EFCs.[4] Furthermore, colleges were not required to present grants to cover any share of the gap between a student’s cost of attending and his federal EFC, and DOJ did not chase after colleges that made no pledge in guaranteeing the gap. If colleges decided to provide grants equivalent to any percentage of the mentioned gap, then it generates a claim for underprivileged pupils that is not available under federal financial aid policy. Definitely, DOJ would not have attempted to stop colleges from doing self-imposed commitments unless the pledge covered the whole gap.  In addition, given greater competition, colleges are expected to cover smaller percentage of the gap as they attempt to evade being flooded by underprivileged and lower quality students. 
  On one hand, the DOJ claims that after Overlap arrangements, colleges are required to give more money to needy students; on the other hand, colleges claim that after Overlap meeting, they would be obligated to give more money to non-needy students and lesser to the needy.

D. The Allegation that the Overlap Meeting Let Colleges Give Too Little Merit Aid
Lastly, DOJ alleged that the Overlap meeting provided the colleges opportunity to distribute less merit aid.
Supposed that the Overlap meetings were efficient in having colleges’ aid-giving behaviors to perfectly correspond to their declared polices and that no-merit-aid policies would have deteriorated without the Overlap arrangements.  Theory implies that this deterioration would increase merit aid relatively within the Overlap colleges, however the striking rise in the mean amount of merit aid would not be in equilibrium. The DOJ’s claim that students attending highly selective colleges would be given much more merit aid was likely because of too literal application of theories on simple firms.
 
V. Did the Overlap Procedures Maximize Colleges’ Objectives or Students’ Welfare?

The exclusion of the Overlap meeting produces a rare opportunity to recognize the core objectives.

A. Banning Merit Aid
Although Overlap colleges, given that they are highly selective, could only hope to save a small amount on grants through banning merit aid, still, they may have been able to save up on a lot of administrative effort and have made connivance on list tuition easier, if they participated in  such collusion.
 Banning merit aid might also help colleges maintain education production technologies wherein contact among pupils of rather similar qualities is important. Without the ban, individual students could not be expected to internalize the impacts of their merit aid on their own education since he would gladly acquire more aid if he could, and prevents other students from receiving merit aid.
If forced to make a choice, a student may go for banning merit aid that makes him acquire a smaller value of aid but also assures him a group of peers wherein they have similar abilities. 

B. Need-Based Aid that Maximizes Colleges’ Welfare at the Expense of Better-Off Students
If colleges have substantial market power, then the need-determined aid may be pure redistribution – it may fulfill societal goals such as assisting the underprivileged students make optimal investments in human capital, although there are no unusual benefits for privileged pupils at the college. In this case, the people who manage colleges get utility from helping needy pupils or accomplishing social objectives, and they are likely to use all the possible rent that could be earned from this benevolent activity. However, under such scenario, colleges could not unilaterally participate in redistribution since pupils who do not value the charity, would at all times accept the proposition of a college that did not participate in charity. Competing colleges rely on collusions on the redistribution formula (and charge privileged students more) in order to participate in redistribution. Collusion would be somewhat difficult to sustain, however if such could be maintained, then the colleges could gain rent provided they are willing to spend it on charity for underprivileged students.

C. Need-Based Aid that Maximizes Students’ Welfare
Suppose need-based aid maximizes student’s wellbeing, albeit individual pupils would like to desert such aid while keeping everyone else to it, then a college offering assistance based on need fulfill social objectives and has unusual benefits for the admitted students. Furthermore, need-based aid draws in capable and varied applicants and allows a college to yield a student body that is more significant than privileged students.   Such aid may also attract contributions from private donors who desire to improve college access.
This regime would need some coordination in order to be maintained, since each student would like to either cheat the aid system, or defect from the regime. If every pupil were to defect, need-based aid systems and packages only attainable thru this regime would seize to exist. Coordination makes it difficult for a student to imply the sense of being poorer than he truly is and make it hard for officer to deviate from the formula he’s supposed to follow. It also keeps colleges from pretending that it is delivering a particular policy when in fact they’re following a different one. This helps in assuring colleges that their policies will gain rewards in relation to their admissions pool and bequests.

VI. Empirical Strategy and Identification Issues

The first test would be on where discontinuation of the meeting led to the financial aids of Overlap colleges to be more akin to those of highly selective colleges that never engaged in the meeting. Second would be on whether the discontinuation of the meeting led to the more gradual increase of tuition (as compared to before) at Overlap colleges. Finally, test is on whether need-based aid is merely redistribution or a method of internalizing externalities.  Unfortunately, such tests are biased toward finding that need-based aid is pure redistribution, since pupils who might be expected to respond to the regime transformation through altering their revealed preferences are unlikely to have fully realized the regime change. In a nutshell, people are likely to respond to how the regime change impacts them personally but under-react to the regime modifications’ more general effects.

A. Empirical Strategy
There are three sections of results: effects on financial aid, on tuition, on students’ revealed preferences. The sections have fundamentally analogous empirical strategy. It is a difference-in-differences approach that depends on two sources of variation:
1. The variation overtime that stems from the antitrust action. A change in the trends of aid, tuition and preferences is expected to be gradual.
2. Only some highly selective colleges took part in the Overlap group. Overlap colleges are expected to have altered their behavior much more than non-Overlap colleges after antitrust action. Also the rather different behaviors of the groups (Overlap and non-Overlap) prior to the antitrust action are expected to have converged after the antitrust action.
If markets were competitive prior and subsequent to the antitrust action, then tuition of colleges that provide alike packages would inevitably act very similarly. On the contrary, if DOJ was right in suspecting collusion on tuition during the Overlap meeting, then Overlap colleges would have chosen (higher) tuition levels that could be backed by non-conspiring colleges. However it is plausible that non-Overlap colleges would have let the Overlap groups to be the price leader and would have behaved as if they were colluding as well.
Moreover, antitrust action significantly raised media attention to tuition of highly selective colleges. All highly selective private colleges endured public relations fallout as a result of the examination and felt pressure to limit tuition hikes due to intense public scrutiny. 

B. The Basic Specification
The strategy is helpful in integrating essential disparities among colleges that are relatively consistent, such as the differences attributable to location, offerings, size, undergraduates’ living arrangements, and the institution’s structure.

C. Possibilities for Omitted Variable Bias
Omitted variable or other simultaneity bias could occur if there are determinants of college behavior which varied inadvertently with the antitrust action and differed more for Overlap and non-Overlap colleges.
There is probability of bias from the financial markets only if the changes affect the Overlap colleges more than non-Overlap colleges. However, this probability is reduced by calculating straightforwardly for various market indicators.
Meanwhile, it is worth mentioning that Overlap colleges do enjoy a degree of likeness to other college thus making coordination feasible.

VII. The Effects of Antitrust Action on Financial Aid

While Overlap colleges reasoned that the meeting assisted them in achieving their objectives of wholly covering need and only allocating aid based on need, the DOJ argued that meeting was pointless for attaining such objectives. In order to verify these claims, there is a need for empirical evidence on college’s policies in grating aid before and after antitrust action.
The National Postsecondary Student Aid Survey (NPSAS) is the source of detailed data on pupils’ financial aid bundles and their family backgrounds and also deals with before and after antitrust action.  The NPSAS was also used to analyze the enrollment of underprivileged students. The Integrated Postsecondary Data Systems (IPEDS) which encompasses colleges’ racial composition of their whole freshman class per year is used to observe the enrollment of minority pupils.  The NPSAS is for the individual-level, while the IPEDS contains institution-level data.

A. The Grant and Aid Functions of Overlap and Control Colleges
For every additional amount in the parent’s income, it appears that among non-Overlap colleges prior to the antitrust action, a student’s grant typically drops by a certain amount. But the amount of this decline is lower compared to the decline in the student’s grant among Overlap colleges. The Overlap colleges’ grants in comparison to the control group seems to be more generous for the highly underprivileged and more elastic with regards to parents’ earnings.
After the antitrust action, control colleges did not seem to have modified their grant functions, as opposed to the Overlap colleges that appeared to have adjusted their grant functions to be less elastic as regards to parents’ income. The decrease of Overlap colleges’ student grant for every 1000 dollars of parent’s income is greater before the antitrust action, compared to after. In general, Overlap colleges do not seem to have adjusted their grants to their most underprivileged pupils.
In examining aid, note that institutional aid does not translate to total aid, which normally consist of federal aid, (and probably state aid and private aid as well). The results suggest that Overlap colleges decreased the total aid by a certain amount for every thousand dollars of parent’s income, this amount of reduction is rather greater compared to the decrease among the control colleges. The control colleges did not significantly modify their aid functions as a reaction to the antitrust action, however, Overlap colleges decreased the elasticity of their functions without substantially altering the aid for their highly underprivileged students. After the antitrust action, aid within Overlap colleges also dropped.

B. Specification Tests
Given that Overlap and control colleges have roughly similar endowment per pupil, they responded similarly to variations in the financial market. The financial market and endowments, albeit declared to be important, do not have an impact on the estimated coefficients. 

C. How Overlap Affected Grants and Aid for Students’ with High SAT Scores
The interpretation for the results focus on verbal SAT scores since they are substantially more discerning than math scores for student applicants of highly selective colleges. It is vital to point out that non-Overlap colleges’ grant and aid functions are more elastic in relation to parents’ income once the college’s predisposition to provide more grants and aid to pupils having high SAT scores is taken into consideration.
Prior to the antitrust action, the Overlap colleges seem to have augmented grants just reasonably for higher SAT scores; afterwards, Overlap colleges seem to have begun offering rather higher grants consequent to higher SAT scores – however, they did not achieve the same level as that of non-Overlap colleges.
Lastly, it seems that before the antitrust actions, the Overlap colleges did not offer supplementary aid for higher SAT scores. In addition, it seems that increases in aid for higher SAT scores explain the majority of the rise in grants for higher SAT scores at Overlap colleges after the antitrust action. Thus the need calculation evidently became fairly more generous for higher scoring pupils.

D. The Family Background of Students who Enroll
The findings indicate no statistically significant time trend in parents’ income at control colleges either before or after antitrust action. However, Overlap colleges’ parents’ real earned income was dropping yearly in the period immediately preceding the antitrust action. In particular, colleges were growingly admitting pupils belonging to lower- and middle-income families. This was, however, reversed after the antitrust action, that the Overlap colleges no longer seem to have a trend in the earned income of parents. (When using the adjusted gross income of parents or the federal formula of expected family contribution, same results were obtained.)
Considering the effect of antitrust on the share of freshmen who are black and Hispanic, not all potential black and Hispanic pupils are needy, but they are disproportionately underprivileged. The overlap colleges argued that extending the need aid to disadvantaged families is a primary instrument for raising the share of able black and Hispanic students in their applicant pools. Before the antitrust action, the Overlap colleges were raising the representation of minorities (black and Hispanic) somewhat faster compared to the control colleges. However these increases were seemingly purged after the antitrust action, when the Overlap and control colleges were both observed to be augmenting their minority representation at slower rates common to control colleges prior to the antitrust action. 

E. Summary
It seems that the removal of the Overlap meeting had an effect on financial aid of participating colleges. In general, grants and aids turned less elastic with regards to parent’s income, somewhat due to privileged parents and pupils having high SAT scores being found to be “needier”. Particularly, there is a considerable flexibility in the definition of need to cover, although small, statistically significant modifications in actual practice. As a result, trends in Overlap colleges towards less privileged, more black and more Hispanic student population were likely to be partly reversed. However it is uncertain as to whether Overlap colleges spend more on financial support as a consequence of the antitrust action. However they offered more grants to well-off pupils and provided them approximately the same amount to the most underprivileged pupils. Nonetheless, without the antitrust action, they probably would have had more underprivileged students.
Outputs suggest that the changes in the Overlap colleges’ behavior even though small are statistically significant and can eventually have considerable impact on the colleges’ student composition. The data implies the impact of Overlap meeting to gradually wear down. Gradualism may probably be the rationale for the antitrust action to seemingly have no indirect impacts on financial aid in non-Overlap colleges.

VIII. The Effects of Antitrust Action on Tuition and Tuition Revenue

The source of data used in this section is from the IPEDS.

A. The Effects of Tuition
Included in the DOJ’s allegations was that the Overlap colleges conspired on tuition at their meeting. However, preliminary analysis on available data does not support the notion that meeting was held as a venue for explicit collusion.
Before the antitrust action, both non-Overlap and Overlap colleges raised their tuition (as well as their total fees) by roughly 4 percent annually. However there are no statistically significant changes in these increased rates after the antitrust action, although the signs of the estimates on the correlations between school year and post-antitrust suggest that there may have been a slower tuition and fee increases among all colleges after the antitrust action that was too small to be statistically significant. 
In a nutshell, the antitrust action does not have a substantial impact on rates of tuition increase.

B. The Effects of Tuition Revenue
DOJ alleged that the Overlap meeting was to allow colleges to give fewer aids to pupils thereby raising the tuition revenue for any given level of tuition.
Preceding antitrust action, the Overlap and control colleges were increasing their tuition revenue fairly less fast than their tuition. The trace of deceleration in the rise of tuition revenue among all colleges was found to be not statistically significant.

C. What has been “Given” to Allow Aid to Change?
There are three possibilities as to how to square the result that the antitrust suit affected financial aid at Overlap colleges but with no modification in their tuition or tuition revenue.
1. The transfers in the composition of student populations toward better-off pupils adjusted for the alterations in the management of individual pupils.
2. Colleges did find funds to raise per-pupil expenses on grants, but the accounting changes made the grants appear as tuition revenue.
3. Colleges reduced college-wide subsidies per pupil (in terms of quality of goods and services, without equivalently reducing tuition).
 The per-pupil costs on grants (like total fees) do not seem to have responded to the antitrust action. The college-wide subsidies, also did not seem to have responded to the antitrust action, but they were observed to increase yearly in both Overlap and control colleges. Thus the evidence indicates that modifications in the composition of student bodies carried the weight of the variations in the financial support. The motive to control the composition of their student population may be why various Overlap colleges backed out from fully need-blind admissions succeeding the antitrust action.

IX. Is Need-Based Aid Just Redistribution

The antitrust action delivers a remarkable opportunity to explore the purpose accomplished by need-based aid among selective colleges.
The findings indicate that middle-earning pupils are less likely to consider an Overlap colleges’ offer following the antitrust action. Moreover, there is no support for or violation of the essential conditions for need-based aid being welfare enhancing regardless of the fact that the (insignificant) increase in the acceptance rates in all Overlap and control colleges were faster among black and Hispanic  private school student who did not acquire any financial support.
 Overall, the available data do not support the idea that need-based is pure redistribution, owing to the fact that Overlap colleges offered richer grants to middle-income aid receivers after the antitrust action but these pupils’ revealed preferences did not shift towards Overlap colleges.

X. Conclusion

There was an observed evidence that antitrust suit did have an effect on financial support in Overlap colleges which led to aid that was increasing less with regards to parent’s income and relatively more responsive to merit. There was no evidence that the Overlap colleges colluded to increase tuition and tuition revenue, and save costs on grants. Students’ revealed preferences prior to and consequent antitrust action do not verify that need-based aid improves the colleges’ students’ welfare through internalizing externalities.  Revealed preferences do contravene conditions required for need-based aid to be pure redistribution.



Source:
Caroline M. Hoxby, “Benevolent Colluders? The Effects of Anti-trust Action on College Financial Aid and Tuition” NBER Working Paper 7754, (June 2000).


[1] Brown University, Columbia University (Columbia College), Cornell University (endowed colleges), Dartmouth College, Harvard University, Massachusetts Institute of Technology, Princeton University, University of Pennsylvania, Yale University, Amherst College, Barnard College, Bowdoin College, Bryn Mawr College, Colby College, Middlebury College, Mount Holyoke College, Smith College, Trinity College, Tufts University, Wellesley College, Wesleyan University, and Williams College.
[2] list tuition – merit scholarship = notional tuition – notional wage
or notional wage = (notional tuition – list tuition) + merit scholarship
[3] For instance, a pupil was enrolled to a certain college wherein the tuition is fully paid by the family (who did not qualify for the aid), and then in the following year the younger sibling gets into a different college, by this time the family would have been “poorer” and is likely to qualify for an aid for the second college. Under the Overlap calculation, the older sibling’s college would not be obliged to re-compute the need for the first child, thus its original obligation was respected.
[4] MIT opposed DOJ’s computations.