Dr.
David Plank
What we are going to do today is report on our research projects
with a particular focus on the systemic impact of choice policies,
how introducing new opportunities for school choice is affecting
the existing education system. Much of what we'll say mirrors
findings from ongoing research elsewhere in the state. As many
of you know well, I think, evaluations of Michigan's charter school
initiative have recently been produced by Western Michigan University
and by Public Sector Consultants, and some of our colleagues at
Michigan State University, including Michael Mintrom from the
political science department, have also been working on choice
policies. But what we will focus on today is the systemic impact
of school choice.
The first section of our presentation
will be done by David Arsen, who is going to talk about some quantitative
data on how choice is affecting the educational system. The first
key question that he will look at is where charter schools are
located and the second is who takes advantage of interdistrict
choice. Gary Sykes will then talk about why the impact of choice
and the competition that choice brings to the education system
has been relatively limited at the school level to this point.
I will then talk about some evidence of innovation in, particularly,
metropolitan education systems in Michigan and conclude with some
observations on present policy issues that have to be addressed
with respect to choice policy and then we'll open up for questions.
I would now like to introduce my
colleague David Arsen.
David Arsen
I am going to use some overheads and speak without the mike,
and I am mostly going to be talking from a handout. What I am
going to do by quickly walking you through a series of tables
is to show you the geographical location of choice in Michigan,
and the participation rates by different types of school districts
for both charter schools and interdistrict choice. So just as
a first step there is Table 1. You can see that the growth of
charter school enrollment was quite rapid. It is increasing at
about 50 percent per year. There is no evident slow down yet.
The state House fiscal agency projects that it will increase another
50 percent next year on the basis of both new schools and increase
enrollment at existing schools. This year there is about 2 percent
of total K-12 enrollment in charter schools. Next year it should
be about 3 percent. Altogether, we are looking at about 30,000
students. The numbers I am about to show you relate to last year
when there were 20,000 students, or about 1.3 percent of total
enrollment. With interdistrict choice, there are fewer students
involved. It's about half. Again, these are numbers from last
year. Last year, there were 20,000 in charter schools, and about
10,000 students in interdistrict choice. That was about .7 percent
of total K-12 enrollment. The percent of districts that accept
nonresident students is slightly more than half. Last year, it
was 45 percent. That is to say, the number of districts that are
accepting nonresident students is growing but is still substantially
less than all of them.
I will now take you through three
tables that essentially describe where the charter schools are
located. This figure breaks up the distribution of 555 school
districts in the state by community type. The classifications
are central city, high income (these would be communities with
median household incomes of greater than $95,000), and other Metropolitan
Statistical Areas. These are all communities that are suburbs
within a metropolitan area but not high income, and then finally
outside MSA and those would be the rural school districts. What
you see is that charter schools are located disproportionately
in central cities. This column here is charter school enrollment
as a percentage of district enrollment. You can see that last
year it was about 3 percent of enrollment on average for those
23 school districts defined as central city. That is a much higher
share than these other sectors. The location quotient is a nice
way of representing that. The location quotient is actually a
ratio of ratios. Basically, for the state it is 1. This implies
that the incidence of charter school enrollment in central cities
is twice the level of the state as a whole. What this shows is
that charter schools are much less likely to locate in high-income
suburbs, and also less likely in the rural areas.
One of the things we're finding
is the charter schools generally do not locate in rural areas
because they lack the critical mass of students and so forth.
They are primarily locating in central city areas.
We have also broken out all of
the school districts in terms of their percentage of black enrollment.
About two-thirds of the school districts in the state have less
than 1 percent black enrollment. You can see that the charter
schools are disproportionately located in districts that have
a high percentage of African American students. In fact, the rate
of charter school enrollment in school districts with greater
than 33 percent black is five times the rate of charter school
enrollment in school districts with less than 1 percent black.
Now notice that this does not mean necessarily -- you cannot infer
from this -- that these students are all black students. They
could be all white students in those highly concentrated black
school districts. What the numbers are saying is that in those
school districts with a high percentage of black students, that
is where we see, disproportionately, charter schools located.
And, finally, just in terms of
MEAP test scores: Where are the charter schools located in relation
to MEAP test scores? I have a breakdown here of the seventh grade
math test and the percent satisfactory on that test. What we see
is that charter schools are much more likely to locate in school
districts with low test scores than in school districts with high
test scores. So on average, charter school location relative to
the student enrollment in school districts is taking place in
central cities, it is taking place in districts with lower incomes
and higher percentage of black enrollment, and with lower MEAP
test scores. Let me just say in passing is that one of the most
interesting things that we've uncovered with regards to charter
schools relates to the cost of students that they are educating.
Some folks have been concerned
about the problem of charter schools creaming, that is to say
taking the best or most talented students away from the traditional
public schools. We don't find much evidence of that. What we do
find evidence of is that they are creaming in terms of the cost
of students, attracting students who are less costly to educate
thereby raising the average per pupil cost of the public school
systems which they leave.
Now interdistrict choice. A key
feature of interdistrict choice is the fact that a school district
has to open itself to the nonresident student. For this kind of
choice to matter, it is not just the demand side but it is also
the school districts opening themselves to students. And they
can regulate the number of seats at which grade level to admit
nonresident students. The first question who is opening themselves
up to choice? Table 6 is a representation of that. Statewide in
97-98 about half the school districts opened themselves up to
nonresident students. School districts that accept nonresident
students are more likely to be central city and rural areas. The
rural districts don't have a lot of activity in terms of charter
schools, but there is a lot of activity in the rural districts
in terms of interdistrict choice. This might have to do with proximity
to school so that there is a relatively high degree of activity
there. Activity is much lower in the high income communities.
Percentage of black students is not a good predictor in terms
of openness. On the other hand, one thing that is a very good
predictor is past enrollment trends. Districts that are declining
in enrollment are much more likely to open to accept nonresident
students. That isn't surprising. Districts that are growing very
rapidly have a hard time creating enough seats just for the resident
students, and are not as likely to open up and accept nonresident
students. This is a powerful predictor and it is highly correlative.
MEAP scores are not a good predicator in this instance. And, finally,
high income communities are less likely to open up to nonresident
students than are low income communities. Not surprising. Whereas
about 60 percent of communities with median incomes of less than
$36,000 accept nonresident students, in communities with median
incomes greater than $76,000, only 23 percent accepted nonresident
students.
Let me move quickly to one bottom
line. We covered a lot of things here but what we've observed
is an upward filtering phenomenon: Students generally moving to
districts with higher socioeconomic and educational standards.
What we are observing here is that interdistrict choice reinforces
enrollment trends over time across districts that existed independent
of interdistrict choice. Those districts that were losing enrollment
are the same ones that losing it as a result of interdistrict
choice. Choice provides a mechanism for people to move to school
districts that they otherwise would have wanted to move to if
they had had the income and could have purchase a house. Here
is one way of characterizing it. Here are several variables and
11,000 students. Take those 11,000 students and for each of those
11,000 students we know the income of the districts that they
live in and the income of the district that they go to, we know
the MEAP test scores of each of the districts, the poverty rate
and so forth. If you look at the table, you'll see that for each
student take you can take the value of this variable from educating
district, where they went to, and subtract the value of their
resident district. What do you get? Well, on average the MEAP
test scores percent satisfactory are about 5 percent higher, reading
3 percent higher, graduation rate higher, drop-out rate higher,
teacher-pupil ratio not much different, spending a little bit
less, salaries a little bit less. But look down here. The districts
the students are moving to are about 12 percentage points less
black, median income about $6,000higher. A measure of poverty,
the percentage of reduced-cost lunches, is about 10 percentage
points less than the districts the students are leaving. And this
just reinforces the comment made a moment ago: The enrollment
in the district they are going to is growing about 3 percent higher
a year than the districts that they left.
Finally, let me just end by saying:
Even though taken statewide, these two choice policies are only
affecting in the neighborhood of 2 or 3 percent of enrollment,
it is very uneven. So we do have a number of school districts
where the impact has been quite dramatic and the districts are
not hard to image. These are school districts that are losing
already in the range of 5 to 15 percent of enrollment. I think
I will end on that for now.
Dr. Gary Sykes
Here's a question: If as a matter of state policy we have
meant to create a quasi-market in education by introducing a set
of new choices into the public school system, would we expect
that the competition that new choice is generating will begin
to have positive effects on the public schools that are currently
in existence? Advocates of competition and market-based solutions
argue that that will happen. So that one argument in favor of
educational choice of various kinds is simply that is a basic
right for families and that it is wrong to deny families the right
to choose their schools. That is simply a value judgment about
which we have argue about in our society. But another kind of
an argument that is more susceptible to empirical test is: If
we introduce choice into a system of monopoly public schools,
are we likely to see a response from the public schools in such
a way that would improve their effectiveness or responsiveness?
When you create a market and there is competition the market theory
suggests that it begins to produce some good effects.
So in another piece of the research
that we've been doing we have begun to test whether or not market
signals are going to have positive effects on regular public schools.
And I will talk about that for just a few minutes. In this case,
I am simply reporting on a telephone survey of a sample of school
principals in the Lansing metropolitan area in which we asked
the principals a series of question about what has happened to
enrollment, what has happened to budgets, whether they felt competitive
pressures from choice, and so on.
Here are our conclusions: By and
large, what we found was that so far (keeping in mind that choice
has only been around for four or five years) the effects of market
signals are relatively weak and uneven. Certainly, when we divided
our data into principals in the Lansing School District versus
others we see much stronger responses from the Lansing principles
because Lansing is one of the districts in the state that has
been most heavily affected. The district's enrollment has really
been affected by both the charter schools and the interdistrict
transfer choice policy. On the whole within this sample, we didn't
find strong responses from principals to the introduction of market
signals of various kinds. I think there are a variety of ways
to explain these findings and we listed some of those on that
slide and I'll just mention each of these factors.
The first factor is limited market
penetration. Right now, competitive pressures only affect 2 percent
of the total population. So one way of arguing for why we haven't
found strong response is simply that there hasn't been very deep
market penetration. The analogy that one of our respondents used
when I was talking to him was the automobile industry. His argument
went like this: When foreign imports accounted for a couple of
percentage points of the total domestic market, the American automobile
companies essentially ignored them. As the percentage of foreign
imports crept up to 10 percent of the market, they began to respond
with things like aggressive marketing, development of niche products
and so on. But they did not change their basic or fundamental
operations. It was only when the foreign competition achieved
market penetration of almost 25 percent of the total domestic
market that the alarm bells really went off among American automobile
makers and you began to get fundamental root and branch reform
of the industry to produce better cars more efficiently at lower
cost. So the argument there is that so far the choice policies
simply haven't achieved sufficient market penetration to cause
the industry, i.e. regular public schools, to respond very strongly.
And by extension, if you continue to believe in the possibilities
of market contribution , you would need to have a more aggressive
set of competitors before you would predict powerful responses
from the public school monopoly. That is one argument to explain
the results.
Another argument is this one: Regular
school principals are already used to the ebb and flow students
in and out of their schools that already exist and that historically
has existed for a long time. Why would an urban school lose students?
Well, there is a series of historical answers to that. People
move their families out of the cities into the suburbs in part
to get better schools. So residential relocation has been a classic
response on the part of families to find better schooling, or
they home school their children, or they pay extra tuition and
then send their children to private schools. Those choice factors
have always operated, and additional increments of new choice
policies on top of the existing choice mechanisms, we find, have
not strengthened the signal from the new choices relative to the
"noise" that already existing in the system where people
are moving around all the time. So from that perspective, since
there is already a substantial amount of choosing that families
have done and the primary choice mechanism has been to move your
place of residence into a district you think the schools are good
and families in America have done that for years, the affects
of new choice policies don't appreciably strengthen the signal
from the marketplace. Consequently, there hasn't been a very strong
response.
A third factor we call preference
versus performance. If we were in situation in the early years
of charter schools in which the charter schools were clearly outperforming
the regular schools on, say, the MEAP. That would be a very distinctive
development because, in part, principals are concerned about how
students in their schools score on the MEAP. But, so far, the
evidence shows that charter schools are not outperforming the
regular public schools. As a matter of fact, they are performing
at a lower level. I hasten to add that that is not a criticism
of the charter schools because in many ways the comparison is
inapt. But here is my point: There aren't strong performance signals
going to regular school principals and instead, we hypothesized,
they interpret the significance of charter schools as sending
preference signals not performance signals. That is to say, charter
schools simply create a new set of options through which parents
can exercise their preferences for certain kinds of schools that
may not be readily available within the public school monopoly.
For example, if you are African American and you want to take
your child to an Afrocentric academy, there aren't any in the
public schools. But there are among charter schools. That's a
preference. In part, what charter schools allow is the flowering
of alternatives that meet parental preferences even if they haven't
yet demonstrated that they constitute superior educational opportunities.
That creates a discounting, psychologically, for principals who
can say: "We're not getting strong signals from our competitors
that they are superior in performance terms. They simply are creating
an option for parents who exercise preferences that within the
public school system because of the rules and constraints we operate
under we can't do." So, again, that would undercut the possibility
of strong market signals.
Finally, this last point. We also
principals that if they lost a lot of students if their budgets
had been trimmed because another kind of a market signal would
be not only losing enrollment, but the budget my go down because
now dollars follow kids. So you lose 10 kids out of your school,
you could lose $6,000 for each student. That gets to be a pretty
sizeable chunk of money. But when you look at the numbers, the
district loses $6,000 per kid, the school loses $69. Why? Because
of the centralization of financing. So, again, if one were to
think that fiscal pressures would send market signals to the school
level, it would provoke strong responses. But there are no fiscal
pressures because of the way the finance system works.
So what we are finding then through
this whole set of factors is that we don't see powerful responses
yet at the school level in response to market or competitive pressures.
Now there is second part of the story. The second part of the
story is that this argument doesn't mean, however, that there
might not be strong pressures at the district level. In fact,
at least in the case of Lansing, we can make a case that pressures
from the market are more likely to be mobilized at the district
level than at the school level. That has interesting implications
and we can talk about that, but I just want to note that when
Chubb and Moe wrote their book and argued in favor of strong market
forces in public education, they also said you had to do it in
a system that had strong site-based management. Control of budgets
had to be at the school level in order to create the proper kind
of marketplace. But what we've done in Michigan is we've created
certain kinds of market pressures but we still have a school finance
system that operates essentially out of the district and is more
centralized. And that is the level from which you see responses
emerge at least in those districts where an aggregate of enrollment
and fiscal effects have begun to be felt and are powerful.
Dr. David Plank
Since I have been the one who has
been signaling my colleagues that they have to abbreviate their
remarks, I'll try to be quick. As Gary suggested at the end of
his section, we are seeing some evidence of innovation particularly
at the school district level. That is still relatively limited
partly because choice policies are relatively new. Charter schools
have only been in operation for five years now, and interdistrict
choice is in its third year. So innovation is beginning to be
discussed, but there is relatively little going on in the ground.
At the same time, as Gary mentioned,
the Lansing school system, for example, was hit hard particularly
by charter schools and subsequently by choice. In the Lansing
school district, there have been a number of competitive responses
that have been imposed on them. One of the first was to introduce
all-day kindergarten because the charter schools came in with
all-day kindergarten. The district is now engaged in a major collaborative
effort with Lansing Community College to start what is called
the Star Institute, which is a vocational training institute.
They have established a community school in the district, the
Wexford Community School, and they have engaged in partnerships
with a wide variety of actors in metropolitan Lansing, including
Michigan State, around efforts to improve the performance of the
Lansing Public Schools.
In other metropolitan areas, there
is likewise some preliminary evidence of change in response to
competitive pressures. In Detroit, one of the first public statements
by the new board that has just been appointed was to call for
the establishment of public schools by the Detroit school district
to provide alternative education for children who might otherwise
be disruptive in the public school system. In Kent County, there
is a task force of superintendents who are looking at the possibility
of charter schools within the Kent County Intermediate School
District to provide specialty educational opportunities for kids
within Kent County, and to compete with some of the charter schools
that have taken a fairly substantial bite out of enrollments at
some school districts. Intermediate school districts around the
state have taken on more entrepreneurial roles in providing services
to local school districts partly on the basis that they will provide
the services more economically than most of the school districts
can provide for themselves. And in Kent County and elsewhere,
there has been encouragement of cooperation and innovation among
member school districts.
Just to conclude, as I can already
sense that there are going to be questions, what kinds of systemic
changes do we see as a consequence of choice? One of the things
that choice does provide is an easier mechanism for social sorting
of the kind that Gary talked about when he talked about preference
versus performance, about parents who prefer a particular kind
of education whether it is African-centered, or whether it is
traditional basic skills. They can now sort themselves out more
readily into schools that provide precisely the kind of schooling
that they want. We see some evidence of market segmentation. Charter
schools particularly are not competing across the board with public
school districts, but instead identifying particular parts of
the market to which to appeal. So you have the Edison Project,
for example, which refers to itself as the Mercedes Benz of charter
schools. They want to be at the high end of the market. They want
to attract the most discerning parents. The New Heritage Academies
have targeted a very different part of the educational marketplace.
Some thing that was not anticipated
in charter school legislation, and that has come to take a very
important role in the public education system as a consequence
of charter schools, is the emergence of management companies.
There are a variety of kinds of management companies that have
emerged. Edison Project is one where they have a particular education
model, which they have both chartered in some schools in Michigan
that they have, in effect, franchised to public school districts.
New Heritage Academy is a very different model, very different
curriculum, and a very different target within the education marketplace.
They have setup academies on the outskirts of a number of metropolitan
areas in Michigan. Others, including the Leona Group, have set
themselves up to provide management services to charter schools.
As both the Western Michigan and the Public Sector Consultant
reports have indicated, management companies have become important
actors in the education system but we know relatively little about
the ways in which they affect the system.
Another point is that the introduction
of choice policy not only affects public schools, it also affects
other actors within the educational system. One of the things
that has happened within in metropolitan school districts is that
a substantial number of the kids who are taking advantage of charter
schools and interdistrict choice were enrolled in private schools.
Private schools are losing enrollment to public education as a
consequence of the introduction of choice policies. As a consequence
is that at least among the private parochial school leaders that
we talked to is that they are coming to an increasingly focused
mission of their religious mission as opposed to a general education
mission. A second consequence is that charter schools that are
independent, that have a unique curricular focus or are run by
what Public Sector Consultants calls "mom and pop operations,"
are under increasing pressure from corporate charter schools.
The corporate charter schools can market more heavily, can provide
services the independents can't provide, and thus the independent
schools are being pressed to the wall by their fellow charter
schools.
One thing that I would like to
say is the that third of the research projects that I spoke about
is funded by a grant from the Michigan Legislature to Michigan
State University, and we are working now on a final report that
should be published in June. In that report, we will reprise the
findings that we have shared with you today and also draw out
some of the implications of those findings for policy action by
the Michigan Legislature. We don't have them ready today, but
we do have some recommendations as to how the policy environments
might be shifted to both accelerate some of the positive things
that are associated with choice and ameliorate some of the problems
that have arisen because of choice.