A school’s efficiency is defined as the
success per dollar spent on students, (with the disparities of incoming
accomplishment of its pupils constant), thus a more productive school yields
higher success in each per student expenditure.
I. Why the
Productivity Consequences of School Choice Matter a Lot
Regardless indirect discussion of school productivity,
previous studies on school choice often overlooked it. Literatures on school
choice focused on distribution questions, such as:
1. Who
practices school choice?
2. Who
selects which school?
3. How
does choice alter the distribution of resources?
Although school choice could theoretically enhance student
achievement through reallocation, such would require that, for every student,
gains from attending a school that is better match beyond the spending enforced
by school choice.
The fundamental logic is that choice provides schools better
incentives to be productive because students favor more productive schools than
less productive schools. Thus, having the same per student expenditures for
both schools, a school would highly likely lose a student to a school that
increase a student’s achievement. Such
process would lead to either two things:
1. the
less productive school is replaced by the more productive replaced; or
2. the
less productive school would raise its productivity and would hence be able to
keep its population of students
In general, improvement in school productivity could be “a rising tide that lifted all boats, and
the gains and losses from reallocation might be nothing more than crests and
valleys on the surface of the much higher water level.”[1]
Thus the reason productivity consequences of school choice
matter is because they likely determine whether all children will gain from
choice. For “rising tide” to occur the concern would be on what efficiency
schools could rationally be expected to achieve.
A. How Much Higher School Productivity
Plausibly Be?
National Assessment of Educational Progress (NAEP) is a
nationally representative measure of student achievement in the United States
which reflects the entire student population’s success and is intended for
evaluation across schools and for long period of time. By calculating NAEP
points per thousand real dollars spent per pupil, the results implied that from
1970-71 to 1998-99 school years, (the CPI-based) productivity fell
significantly greater than 50 percent and 70 percent in terms of math and
reading, respectively. If schools went back to 1970-71 reading productivity and
1972-73 math productivity, the mean American pupil would attain a score that
lesser than 10 percent of American students presently achieve.
Meanwhile in consideration of the demographic changes, it
may be that schools were not losing productivity, instead they were merely in
operation with students belonging to a family with poor background.
The standard approach to address such issue is to:
1. Identify
the impact of attributes such as African-American, Hispanic, single-parent
family, family income, and so forth;
2. Evaluate
what would be the 1998-99 achievement provided that the student population were
equal to the 1970-71 student population;
3. Identify
what would be 1998-99 productivity given that student population was the same
as that of 1970-71.
It seems that taking such method into consideration, but
adjusting for demographic increased the decline in productivity by
approximately 3 percentage points. Increase in the decline is due to smaller
fraction of student having high school or college graduate parents (in
1971-1972) than the latter years. In addition, pupils having parents who
dropped out of high school seems to attain lower scores on the NAEP exam than
their peers. Meanwhile, since 1971-72, the fractions of African-American and
Hispanic pupils have increased, and they tend to score worse on NAEP exam in
contrast to non-Hispanic white students. On the other hand, the impact of the
changing racial composition is overshadowed by the impact of the changes in the
parent’s education and income.
Another factor to consider for the drop in school
productivity aside from demographic changes is probably the changes in career
opportunities for women. Explicitly, over the period 1970 to 1999, non-teaching
opportunities for women were opening up, thus making it more and more costly
for schools to hire a female with a certain level of skills.
Thus by adjusting for salary of women having advanced
degrees, the fall in productivity ranging from 39.1 to 57.6 percent, though
small (compared to CPI-based decline) is still substantial.
Findings imply that the main cause of the drop in
productivity is not the changing student attributes nor female career
opportunities, but the school conduct. Thus policies that improve school
conduct could likely trigger large boosts in productivity. Stating this,
however, is not enough. Whether choice actually causes schools to increase
productivity must be investigated.
B. How School Productivity Affects
American Industry and Growth
United States’ comparable advantage in making goods and
services that intensively use educated labor is because America always had a
relatively plenty of educated labor – thus United States has constantly been
able to generate education rather cheaply. America’s “new economy” products
such as microprocessors, software, and knowledge services, are some of the most
human capital intensive goods in the world. Moreover, from basic trade theory
that human capital intensive economy is based on the Americans capability to
somehow inexpensively generate education in its population. Although Americans
can import human capital, it cannot be a source of comparative advantage in the
mid-to long-run.
Therefore, if Americans wish to maintain their growing
economy that is concentrated on human capital intensive goods, they must
concern themselves on the quickly declining productivity in their schools – as
dropping school productivity leads to America’s having a somewhat more
expensive human capital, which then leads to a loss of comparative advantage in
human capital intensive goods.
This implies that the effect of choice on school
productivity does not only have an overwhelming impact on the allocation
effects of choice on achievement but it also has a wide connotations for the
macroeconomy, for trade, and for Americans’ occupations.
II. How
Productivity Fits into the School Choice Literature
Previous literatures on school choice have somehow
disregarded the productivity implications of choice. This neglect is due to the
foundations of the theoretical literature. School choice models have developed
out of local public goods provision models, which have conventionally concentrated
solely on allocation issues, for instance what local public good is offered and
how a person’s local public good choice affects others. School choice
literature has been influenced by such focus, however, previous literature must
not dictate neglect of productivity.
Examples of the importance of productivity effects when
competition is introduced into a market:
1. Health
care: after the passing of legislation in 1980s-1990s, managed care
organizations were permitted to compete. Effects of competition were not just
on the allocation of health care, but more remarkably on productivity, which
was beyond the expectations of the managed care supporters. Productivity seems
to develop in a more competitive condition, partly because competition
encouraged providers to take on efficiency enhancing technology and to put off
conduct created rents.
2. Trucking
services: deregulation in the 1970s led to the remarkable growth in trucking
competition. For the same amount, a trucking client is provided faster and more
specialized service than before.
3. Parcel
services: competition enhanced productivity not only because the private
companies had higher productivity and yield growth than the United States
Postal Service, but because it also made the United States Postal Service to
significantly boost its own productivity.
It is rather peculiar that school productivity is neglected
in the school choice literature since there is growing interest among
economists in the productivity of not-for-profit, semi-public and regulated organizations.
It is worth mentioning that one type of school-related study
does contain considerable evidence on productivity, though productivity is
hardly ever cited and calculations for which are never made. Such research
compares students’ yields in public and private schools and tries to get rid of
selection bias, which may occur from the fact that the difference of students
who “self-select” into private schools compared to students who stays in public
schools may be unobservable.
Reformulation of the research as a comparison of public and
private school productivity is due to the continuous attempt to evaluate
accomplishments and hold steady the quality of student inputs. However, other
inputs, such as spending, are not constant and this fact is often disregarded.
Particularly, the common private school per pupil spending in the United States
is roughly 60 percent as much as that of the common public school, on the other
hand, private school spending is much more erratic than public school spending
thus minimum (maximum) private school expenditure is below (above) the minimum
(maximum) public school expenditure. Hence, although researchers are able to
observe public and private schools producing similar yields, it would likely be
factual that private schools are substantially more productive as their mean
costs are lower. In order to make a precise comparison of public and private
school yield, productivity calculation for each school must be made in order to
consider the differences in the spending distribution and these calculations
are to be compared for students having similar backgrounds.
III. Why Should
Choice Affect School Productivity?
This is mainly a query on:
1. What
is maximized by schooling producer; and
2. What
the schooling production function is like?
The answer this question varies for different types of
schooling producers: for-profit firms, not-for-profit private schools, charter
schools, and regular public schools.
For all types of producers, one assumption is maintained:
For any given cost, parents will select the school that produces the schooling
that they deem most valuable.
A. A
For-Profit School Producer that Takes Up Charter School Contracts
A for-profit firm that starts a charter school is somewhat
typical of Edison Schools and may become a general model if charter school
programs were more extensively enacted.
The law sets the fees imposed by schools, and parents are not able to
“top up” the cost. In addition, it is also assumed that the school’s plan to
accept or reject pupils cannot be included in the school’s profit maximization
strategy.
Then the concern of the school is maximizing the difference
between revenues (the product of the fixed fee and the number of pupils who
enroll) and costs (the product of per-student costs and the number of pupils
who enroll).[2]
The per-student cost is an increasing function of the quality that the school
offers, the labor that it employs, and other inputs it utilizes. Moreover it is
also a function of the salary rate for the employees and the going price for
other school inputs. It is assumed that per-student costs are similar in spite
of the school’s size – however such assumption may unlikely be true.
Meanwhile, the number
of enrollment is an increasing function of quality since parents chooses the school
that provides the highest quality for a certain price.
Now, assume that the school shares similarly in enrollment
if it presents precisely the same quality as another school.
In such conditions, the best option for school is to
maximize the subject quality given the restriction that its per-student
spending must not go beyond the charter school fee. In other words, the school
should maximize its productivity for a certain cost to prevent another school
from taking in all the students in the area. Thus it is not surprising that
unproductive schools will be eliminated from the market.
The school can maximize the number of pupil on whom it gets
its slim profit by providing the highest possible quality that the charter
school fee can bear. Managers of for-profit schooling firms consider that there
are economies of scale in schooling since a firm can pay lower costs for its
inputs if it consolidates purchasing, curricular research and development, and
information processing over various schools. Given economies of scale, huge
firms are likely able to earn economic profits (profits surpass the earnings
needed to pay the cost of capital) in local markets where they compete with
other schools that are too small to exploit the economies of scale.
B. A
For-Profit School that Takes Up Vouchers
A for-profit school firm that takes voucher students is
relatively comparable to what was discussed above, apart from, however, the
consideration that parents are implicitly permitted to “top up” a voucher with
extra tuition payments from their own funds.
If not, suppose that the case is similar: the school haves
to allow voucher applicants on a random bias depending on their willingness to
pay the school’s fees with a combination of the voucher and additional tuition
costs.
Since school set its fees, the concern becomes more
complicated, as the school must take into consideration the fact that a higher
fee implies not just larger revenue per pupil who enrolls but also fewer
enrollment – since higher fee puts off enrollment for any given quality level
the school provides. Thus for any given fee, the school should maximize the
quality it produces provided that the costs are less than or equal to the given
fee so as not to lose its entire enrollment to another school that provides
higher quality for the same costs. The new equation allows for the economies of
scales.[3]
Hence, a school aiming for a “better” but smaller niche of parents who are
eager to take on higher fees for higher quality, must consider the loss of
economies of scale and consequently, increase in its costs.
C. A
Non-Profit School that Takes Up Charter School Contracts
Not-for-profit firms are the huge majority of school
producers that engage in charter school contracts or voucher pupils. The
primary distinction between a non-profit and a for-profit institution is the
surplus distribution as the latter allocates profits to its owners (private
owners or shareholders). Thus, it is logical to assume basic profit
maximization since owners instantly benefit from profits. If a non-profit
school has surplus (difference of revenue and cost), it cannot directly, by any
means pay them to anyone. On the other hand, surplus can be considered
significant since it can be utilized in various ways:
1. It can
be used to allow working environment satisfying for the school’s staff
regardless if such conditions do not add to productivity.
2. It
lets a school follow social goals that its staff consider valuable such as
exploring teaching methods, creating new curricula, and so forth.
Some of the things to consider on the surplus distributions:
1. They
are almost always ineffective compared to cash distribution as portions of the
surplus are lost in the process converting it into goods or services that the
employee values. Thus, a school worker receives weaker incentives than what
they would receive if they could be given cash incentives.
2. Though
it is fairly easy to allocate a non-profit school’s surplus to its employee, it
is difficult to legally issue it to a particular owner. Hence, a school is less
motivated to expand merely to increase the absolute surplus size – as schools
expand surplus will rise, however, the number of staff over whom the surplus is
distributed will also increase. This is contrast to the for-profit condition
where owners have motivation to increase their schooling production provided
they can get some positive surplus per additional enrollee.
Hence the under the new maximization problem, although the
school’s motivations to increase enrollment are weaker compared to a for-profit
school, its motivations to maximize productivity are strong. The school is
overwhelmed by its competitors if it does not generate the maximum quality
possible given that its costs do not exceed the fixed charter school fee. [4]
Comments not-for-profit school’s maximization problem:
1. Provided
there are economies of scale, school will have greater incentives to expand
enrollment.
2. If the
school produce surplus and use it to buy staff rewards that appear to be inputs
– although this do not add to outcomes that parents value – one’s measure of
the non-profit school productivity may somewhat understate its true
productivity. The understatement will be slim since competition among
non-profit schools will stir surplus to zero.
D. A
Non-Profit School that Takes Up Vouchers
The not-for-profit voucher school should maximize
productivity so it does not to lose its enrollment to equivalent schools that
provide higher quality for the same price. The only problem is that the school
must select its fee and quality at the same time, and the only stipulation is
that the school has weaker motivations to increase enrollment than a for-profit
voucher school.[5]
E. A
Summary for Fee-Based Schools (For-Profit and Non-Profit)
For all the cases, the school’s revenue is gained from
student fees. The fee basis is vital since it implies that parents’ choices
decide whether a school is practical or not. If a school’s pupils are drawn
away by a competing school that imposes the same prices, logically, the school
must enhance its productivity. Furthermore, even though only for-profit
schooling firm will have greater motivations to go into new markets and expand
new enrollment, both not-for-profits and for-profits have reasons to maximize
productivity.
It is often questioned whether there will be an elastic
charter or voucher schools supply. This is vital for non-profits in which
obvious incentives to grow is absent as they only wish to earn a small surplus
on additional students. If economies of scale are present, they suggest that
both for-profits and non-profits must have elastic supply once they are in
business. On the other side, there are some factors that may operate like
diseconomies of scale, for example, a charismatic principal managing a large
school may likely lackluster due to little direct contact with students.
Buildings are often taken as a possible factor that would
limit the charter or voucher schools supply elasticity. However, such seems to
be a short-run phenomenon that is primarily observed in start-up of new charter
or voucher programs. The total number of student enrollments does not increase
merely because of an entry of a new school, hence, the presence of charter or
voucher school competition does not need much of a net increase in school
building. So as students transfer to a more productive school from less
productive one, buildings should be sold (purchased) by the shrinking or
exiting schools (expanding or entering schools).
Still schools are rather undividable and although parts of
school buildings are often sold or leased to separate schools, only particular
portion of a building will generally create a feasible school. Rationally,
competition calls for a slight increase in the total stock of school buildings,
to provide more flexibility as parents’ capacity to select makes enrollment
more variable.
Moreover, some factors (economies of scale) imply that
school supply will be extremely elastic, while other factors (diseconomies of
scale) imply that school supply will be less elastic.
F. Competition
and the Productivity of Regular Public Schools
A typical public school in competition with a charter or
voucher school, with the latter’s fee coming directly from its budget, then the
former is fee-based at the margin and will have marginal incentives to be
productive. The effectiveness of these marginal incentives relies on the scale
of the fee or voucher as some vouchers or fees are so small in relation to
regular per-student costs that they provide public school perverse motivations
to push students away – and for each of these students, their per pupil
spending increases non-negligibly. Moreover, public school employee might be
able to gain larger surplus by driving students away than trying to draw them
in. Such twisted scenarios can be evaded by setting an adequately high voucher
or charter school fee.
However, a regular public school not facing competition from
a charter or voucher school and is also not fee-based at the margin, still has
incentives to be productive. A public school is funded by local property taxes
and experiences a high level of traditional choice among public school
districts – which occurs when parents select a school district by selecting a
residence. This traditional choice is certainly the most insidious and
significant form of choice in American elementary and secondary schooling. So
that this form of choice provides schools incentives to be productive, it is
crucial that parents pick among districts that are fiscally independent.
Normally, United States public school districts have
revenues that rely heavily on local property taxes, hence parents, in a
metropolitan area, able to select among a large number of districts, will be
inclined to favor districts that produce higher achievement for a certain local
property tax liability or, equivalently, have lower local tax liability for a
particular achievement level. In other words, parents are apt to prefer
districts having high productivity, thus they tend to avoid school district
with declining productivity which would result to decrease in the demand for its
houses which in turn bring down the district’s property prices, which
consequently, bring down the school’s budget, which depends on property tax
revenues. The administrator will be urged to increase productivity, either via
maintaining achievement given the lessening budget or through substantially
increasing achievement, making the district appealing to homebuyers once more.
It was found that the households’ observations of
achievement and tax liabilities are “universalized” by means of the housing market
in order that every family’s house price shifts in a certain way as to provide
schools incentives to be productive. This productivity-inducing mechanism[6]
can be long-lasting since it relies on decentralized choices, as opposed to
centralized reward systems which are likely to be unsound since state
governments cannot, ex post, credibly stick to systems that lessen the amount
of money distributed to failing school districts.
IV. Finding
Evidence on How Choice Affects School Productivity
Some problems may arise in any analysis of how choice
affects productivity:
A. The Endogenous Availability of
Choice Options
One problem is the fact that choice options do not occur
randomly, but are often a reaction to school conduct, particularly, when people
are displeased with a certain conduct of a school, they try to generate (or
maintain their access to existing) alternative schools. This occurrence can be
observed relative to the production (or maintenance) of private schools,
charter schools, and voucher programs. Parents, who are frustrated over the
terrible public schools in a given area, are willing to put some effort or
devote some money to attain alternative schooling. This produces an area
wherein private schooling is widespread since the public schools are terrible.
Furthermore, the presence of choice options due to bad
public schools is an issue that is also present in traditional choice among
public school districts. Voters refuse district integration in areas with one
or more districts having bad productivity; and expectedly, voters support
integration provided all the districts have good productivity, this in order to
take advantage of economies of scale. However, since districts having bad
productivity, sub-areas are eager to break away to form another district, while
districts with good productivity, withdrawals do not occur , hence, areas
having numerous districts often have one or more districts with bad
productivity.
Endogenous school choice in areas where public schools are
bad produces bias if a researcher plainly estimates the choice effect on
productivity. Because schools having bad productivity stimulate choice
creation, and it might seem as if choice induce poor productivity. Such bias
can be avoided by:
1. Evaluating
the same school district before and after a choice reform
- This
normally creates “differences-in-differences”
strategies.
2. Finding
a source of disparity in the choice availability that is not associated with
the underlying factors of poor school productivity.
- This
normally creates instrumental variables strategies.
B. Unobserved Differences in Student
Inputs that Appear to be Differences in Productivity
Meanwhile, in measuring a school’s productivity, the
differences in pupil inputs must be fully considered as to avoid identifying a
run of the mill school as highly productive just because it has good student
inputs – achievement is high even though it adds minuscule learning on top of
what the students learn outside school process. However, there is difficulty in
observing, measuring and controlling all student inputs such as enthusiasm and
natural skill.
There are three ways to deal with this concern:
1. Comparing
schools having high choice-based incentives such as voucher or charter schools with
low choice-based incentives like large public school district that leads a
metropolitan area, in which it must be guaranteed that an arbitrary mechanism,
not correlated with unobserved motivation/skills, allocates pupils to schools.
Given such mechanism, schools will have an equivalent distribution of
unobserved motivation/ability, and the disparity in achievement per dollar
spent will precisely reflect true variations in productivity.
2. Comparing
students’ accomplishments from an environment having little or no choice to an
environment having lot of choice, provided that pupils cannot select the
environment to go into, such method produces good estimates.
3. Evaluating
the achievement of pupil who is unlikely to gain from choice except if all students gain. For instance, in
comparing school productivity before and after a private school becomes
available, it was observed that the private school attracts students who are
high achievers formerly in the public schools, hence, the measured productivity
at the public schools before and after the private school’s opening can be
estimated, knowing that public schools’ measured productivity is expected to
increase only if all students benefit from the private school choice
availability.
C. Measuring Productivity
Productivity is achievement per dollar exhausted in a school
and issues on productivity measurement are mostly linked to measuring
achievement. Using gauges of achievement like grades, having diverse
interpretations in various schools and times must be avoided. In addition,
using standardized tests administered only to a small number of pupils, for
instance, Scholastic Aptitude Test (SAT1) or American College Test (ACT), must
be avoided because it generates self-selection bias that is impossible to fix
without the use of other standardized tests that are to be taken by the entire
student population. However, if such a population wide standardized test
exists, then it should be used rather than the SAT1 or ACT. Moreover, given the
broad standardized test, would raise another concern on what score to be used
(reading, math, science, elementary, secondary, and so on).
Measuring per student expenditure face few concerns provided
similar definition is used for all schools.
V. The Effect of
Traditional Forms of School Choice on Productivity
Through mechanisms such as ability to select among public
school districts and to select private schools, parents have traditionally
practiced some choice on schooling of their children. Such traditional forms of
choice are helpful in estimating the impact of choice on productivity,
particularly since, the availability of traditional choice mechanisms largely
differs across United States metropolitan areas. Some areas hold several
independent school districts and plenty of inexpensive private schools, while
other areas are fully monopolized by one school district or there is almost no
available private schooling.
Additionally, traditional choice provides significant
evidence on productivity that is somewhat unavailable and shows empirical
strategies for determining the impacts of traditional choice.
A. Traditional Inter-District Choice
The first traditional form of choice takes place when
parents select among independent public school districts through selecting a
residence, and the extent to which this can be used by parents relies on the
number, scale, and housing patterns of districts in the area of the parents’
occupation. A regular metropolitan area has an amount of choice that
corresponds to containing four equal-sized school districts (or a bigger number
of less uniformly sized districts). On the other hand, people having
occupations in rural areas normally have just one or two school districts to choose
from. When studying traditional inter-district choice, it is better to focus on
metropolitan areas as to evade a much-choice/little-choice evaluation that
primarily mirrors urban/rural disparities in school productivity,
In order for traditional form of choice to be helpful in
determining productivity effects of choice, it is crucial for parents to pick
among fiscally and legally independent districts. Since the mechanism above
does not function, if a district is entirely reliant on state revenue or is
taken safe from consequences linked with incapacity to draw parents.
Intra-district choice among schools does not offer valuable evidence on
productivity effects since by definition, the schools in a district are
fiscally dependent on one another.
Furthermore, a fine index of inter-district choice is the
likelihood that, in an arbitrary encounter, two pupils in the metropolitan area
would be admitted in different school districts. If only one district is
present, the probability would be zero; and if there are several districts, the
probability would approach one (greater than 0.95). [7]
B. Traditional Choice of Private
Schools
The second means in which parents exercise traditional form
of choice is through admitting their children into private schools. America
private schools are not subsidized by public funds hence, parents can afford
private school only if they can pay tuition at the same time, pay taxes that
sustain local public schools. Partially, as result, only 12 percent of American
students are from private schools. Moreover, 85 percent of private school
students are enrolled in a school with religious affiliation, having tuition
that varies from a coupon amount to above 10,000 dollars; while, the remaining
15 private attend private schools having no religious affiliation, which charge
tuition of 5,000 dollars or higher. The average Catholic school student in the
United States is charged between 1,200 and 2,700 dollars for tuition.
The main feature of American private schools is that tuition
is normally subsidized by revenues from donations or an endowment. Schools that
serve low-income pupils have bigger fraction of schooling cost covered by
subsidies, however even fairly costly private schools impose subsidized
tuition.
Meanwhile, given a certain level of quality, the quantity of
private school places that are available at a particular tuition (or the
private schooling supply) significantly differs among United States
metropolitan areas. Therefore the extent to which parents can exercise choice
between public and private schools changes among metropolitan areas.
C. Why Evidence from the Traditional
Forms of Choice is Necessary
Evidence from the traditional forms of choice is essential
since it can disclose the long-term, general equilibrium impacts of choice,
while evidence from recent reforms cannot.
In the short term, an administrator trying to raise the
school’s productivity in respond to competition has merely certain options:
1. Make
the staff to work harder;
2. Do
away with unproductive employee and programs;
3. Distribute
resources away from non-achievement oriented objectives such as building self
esteem, and into achievement oriented ones such as math, reading, and so forth.
4. In the
medium term, the school can be made more effective through the renegotiation of
the teacher contract.
By exercising all of these options, the administrator may be
able to substantially enhance the school productivity.
Nonetheless, productivity can be affected by choice through
a range of long-term, general equilibrium mechanisms that are not directly
available to an administrator. The financial stress of choice may command
higher salaries for teachers whose teachings increase achievement and draw in
parents. In turn, it may attract people into teaching or maintain existing
teachers who may as well follow other professions. Certainly, it could alter
the entire reward structure in teaching and thus transform the occupation.
1. The
importance of pulling parents in might induce schools to release more
information about their achievement and thus could gradually turn parents into
better “consumers.” Since parents’ choices are more significant when schools
are funded by fees they control; choice can make schools more open to parent
participation.
2. The
significance of producing competitive outputs in relation to other schools may
drive schools to identify and discard academic techniques and curricula that
are ineffective.
In the long-term, the scale and existence of schools is
affected by choice:
1. Choice
increases and decreases districts’ enrollments;
2. Choice
makes private schools enter and exit.
D. The Effect of Traditional
Inter-District Choice on School Productivity
Provided with the good measure of the degree of
inter-district choice in a metropolitan area, the concern in particular is that
districts merge with productive districts but withdraw from unproductive
districts. Hence to get unbiased estimates, geographic or historical factors
(that raise a metropolitan area's inclination to hold many independent
districts but have no straightforward impact on modern public school conduct)
are needed.
The effects of inter-district choice is estimated using
regressions in which the dependent variable is either achievement or per
student costs, with the choice index to be the main independent variable. The
primary disparity in the regression is found in the metropolitan area level;
however there is a control placed on a wide range of background variables
(household income effect, parents' educational background, family size, race,
region, metropolitan area size, and the local population's income, racial
composition, educational attainment, and so forth) that may also possibly
affect schools or students.
The key results of these regressions illustrate that
inter-district choice has a positive, statistically significant effect on
productivity with at least 95 percent confidence. It shows that the schools are
able to simultaneously have significantly higher accomplishment and
significantly lower costs provided their productivity is considerably higher.
E. The Effect of Traditional Private
School Choice on School Productivity
To measure the impact of changing private school competition
for public schools, factors that have an effect on the private schooling
supply, but have no direct impact on achievement are needed. Such factors
comprise of historical differences in metropolitan areas’ religious composition
since religious groups left donations that make disparities in the amount of
non-tuition revenue enjoyed by private schools.
As long as there is a control for the present religious
composition of metropolitan areas (which may have an effect on the demand for
private schooling), these historical religious population densities must
majorly affect the schooling supply and must have little or no direct impact on
the of public school students’ accomplishments.
Private school choice effects is estimated using regressions
wherein either achievement or per student costs is the dependent variable, and
the primary independent variable is the fraction of metropolitan area students
in private schools. There is control for the same background variables that is
used for inter-district choice (mentioned above).
The result shows that private school choice has a positive,
statistically significant impact on public schools’ productivity. Also,
traditional private school choice does not affect public school expenditures,
since rise in the private school choice availability draws some students away
from the public schools, increasing per student costs through the decline in
the number of pupils served yet lessening per student cost through the
reduction in voters who will support higher public school expenditure. Moreover
the substantial effect on public schools’ productivity by private school arises
purely through an effect on achievement: public schools’ per student costs does
not change, but their achievement is higher.[8]
F. Discussion of the Effects of
Traditional Forms of School Choice
If every school were to face high degree of inter-district
choice and private school choice, in contrast to zero inter-district choice and
fairly low private school choice, there would be a 28 percent step up in
American school productivity
However, traditional forms of choice offer slightly feeble
incentives as opposed to choice reforms such as vouchers and charter schools.
Furthermore, many underprivileged families cannot use traditional forms of choice,
for instance:
1. A
family can only select among districts if it has sufficient funds to live in a
range of areas;
2. Traditional
private school choice can be practiced only if the family can pay the tuition.
And so, regardless if every United States metropolitan area
had the maximum level of traditional forms of choice, underprivileged families
are most likely to be left with rather unproductive schools.
VI. The Effect of
Recent Choice Reforms on School Productivity
Current choice reforms can only partly answer questions on
competition’s effect on productivity. The reforms’ most modern vintage implies
that observers are unlikely to see major changes in the supply of schools. In
addition, there can be difference between short-term and long-term reactions to
choice. Take for example, a typical public school, (that had shown low
productivity for years), that has become the target of voucher or charter
school competition. That school may possibly, under pressure:
1. Make
striking productivity gains in the short run;
2. Hastily
discard unsuccessful instructional programs or staff;
3. Quickly
move resources into core instructional programs in reading, language, math,
history, and science.
However, the productivity increase rate may possible slow
down after the first few years as good policy changes become less evident. Then
again, even a school that is increasing its productivity might seem to have
productivity losses in the short-run if it subjected to adjustment expenses
when it makes changes.
In order to learn from recent choice reforms, few principles
must be followed:
1. It is
essential to study the productivity reactions of regular public schools that
are recently facing competition, since their reactions are in much more
hesitation than the productivity of choice schools. An unproductive choice
school is less apt in entering or even surviving, however opponents of school
choice doubt whether typical public schools even have the knowledge or tools
needed to enhance their productivity.
2. There
is a need to focus on the productivity reactions of regular public schools that
experience non-negligible incentives due to a choice reform. This instantly
restricts the investigation to a few choice reforms that meet the following
requirements:
a. There
is a likelihood that no less than five percent of regular public enrollment
could attend choice schools,
b. The
typical public schools lose some money (may not be the full per-student cost)
when a pupil enrolls to a choice school,
c. The
reform has been in existence for a few years.
Three reforms that satisfy these basic conditions are:
a. School
vouchers in Milwaukee,
b. Charter
schools in Michigan, and
c. Charter
schools in Arizona.
A. The Effect of Vouchers on
Achievement in Milwaukee Public Schools
The results imply that public schools have a strong,
positive productivity reaction to competition from vouchers; hence schools
facing the most possible competition from vouchers had the best productivity
reaction. Furthermore, schools most treated to competition had remarkable
productivity enhancements. On the one hand, such productivity growth bursts may
actually decelerate after a few more years of competition. Meanwhile, the
competition productivity effects may be understated since the school’s control
group was a somewhat unfair comparison group having fewer underprivileged and
minority pupils.
B. The Effect of Charter Schools on
Achievement in Michigan Public Schools
It was observed that as a response, public schools that
experience charter competition increased their productivity and achievement, not
just beyond their own prior performance but also improving with respect to
other Michigan schools not facing charter competition. The enhancements in
productivity and achievement seem to arise once charter competition arrives at
a critical level that corresponds with the enrollment at which charter schools’
admitting pupils would be easily noticeable (this must not be confused with
regular fluctuations in enrollment). In addition, the rise in productivity and
achievement is greater and more accurately estimated in fourth grade, possibly
because elementary schools experience more competition from charter schools
than middle schools did.
C. The Effect of Charter Schools on
Achievement in Arizona Public Schools
The results provide conclusions generally analogous to that
of Michigan’s. Charter competition concentrated on public schools that
initially had achievement and productivity growth below average, but charter
competition caused public schools to enhance their productivity and
achievement. The improvements are in relation to the schools’ own previous
performance and also relative to achievements over the same period, by schools
that are not faced with charter competition.
D. Discussion of the Effects of Recent
School Choice Reform
All three forms of choice did increase productivity. If all
schools were to experience productivity growth rates similar to those in
Milwaukee’s most treated schools, American schools might revert to their
1970-71 productivity levels in under a decade.
On one hand, the productivity growth bursts observed in Milwaukee may
eventually relax to a lower growth level. On the other hand, several long-term,
general equilibrium choice impacts are still not in operation.
Furthermore, if the student is enjoying the success growth
rates that pupils in the most treated schools are currently enjoying, that
student will “grow out of” the bad distribution effects within a few years,
hence, that student will be at an advantage for having experienced vouchers
within 5 years of the voucher program.
VII. Conclusions
This paper presented evidence that implies that school
choice productivity effects, must be given importance not only because they
potentially alleviate the tensions created by the distribution effects of
choice but also because American schools are in a productivity crisis. In order
to return the schools to their 1970 productivity levels, policies that improve
American schools’ productivity are greatly needed.
Meanwhile, schools that experiences choice-driven incentives
can be encouraged to increase their productivity. In addition, traditional
forms of choice also raise school productivity, specifically, if all schools in
the United States are faced with high levels of the traditional forms of
choice, school productivity could be as much as 28 percent higher than it is
currently.
Moreover, typical public schools raised their productivity
when presented with competition. In all cases the typical public schools
increased productivity by raising achievement, and not by reducing spending
while maintaining achievement. Such achievement-oriented reaction may, somewhat
be related to the actual reforms’ attributes.
Overall, it appears safe to conclude that studies that
disregard the productivity impacts of choice are prone to be deceptive. The main
effects of choice may be improvements in productivity.
Source:
Caroline M. Hoxby, “School
Choice and School Productivity (or Could School Choice be a Tide that Lifts All
Boats?” NBER Working Paper 8873, (April
2002).
0 comments:
Post a Comment