THE IMPACT OF CHARTER SCHOOLS ON CATHOLIC SCHOOLS: A COMPARISON OF PROGRAMS IN ARIZONA AND MICHIGAN

INTRODUCTION The American public school system is vast. More than 15,000 school districts govern over 95,000 schools employing over 6 million people serving 50 million students and spending $500 billion per year (Hess & Finn, 2007). Public schooling, in short, is a colossus casting a very long shadow. Major reform efforts within the public education system will inevitably influence the private school sector, sometimes profoundly so. Even as by far the nation’s largest system of private schooling, at 2.3 million students (McDonald, 2006) the size of the Catholic school system pales in comparison. Nevertheless, Catholic schools have a proud tradition of outperforming public schools, in particular with disadvantaged students. In vast swaths of urban America today, Catholic schools remain the highest performing schools available to inner-city youth. Americans have been attempting to reform and improve public schools since the launch of Sputnik in the 1950s. Education reformers have tried everything from increased spending to progressive instruction techniques to expanded early childhood education, smaller class sizes, and even open classrooms. Through it all, the most reliable test scores available (Perie, Moran, & Lutkus, 2005) have remained stubbornly flat since the early 1970s. With far greater spending per pupil and without improved learning, the productivity of spending in the public education system plummeted. In addition, there has been a growing recognition of the inequities of the public edu-


INTRODUCTION
T he American public school system is vast.More than 15,000 school dis- tricts govern over 95,000 schools employing over 6 million people serving 50 million students and spending $500 billion per year (Hess & Finn, 2007).Public schooling, in short, is a colossus casting a very long shadow.Major reform efforts within the public education system will inevitably influence the private school sector, sometimes profoundly so.
Even as by far the nation's largest system of private schooling, at 2.3 million students (McDonald, 2006) the size of the Catholic school system pales in comparison.Nevertheless, Catholic schools have a proud tradition of outperforming public schools, in particular with disadvantaged students.In vast swaths of urban America today, Catholic schools remain the highest performing schools available to inner-city youth.
Americans have been attempting to reform and improve public schools since the launch of Sputnik in the 1950s.Education reformers have tried everything from increased spending to progressive instruction techniques to expanded early childhood education, smaller class sizes, and even open classrooms.Through it all, the most reliable test scores available (Perie, Moran, & Lutkus, 2005) have remained stubbornly flat since the early 1970s.
With far greater spending per pupil and without improved learning, the productivity of spending in the public education system plummeted.In addition, there has been a growing recognition of the inequities of the public edu-cation system.In recent years, reformers have focused on addressing the racial achievement gap.
Parental choice in education has emerged as a major education reform movement since 1990.Choice programs within the public school sphere, in the form of open-enrollment programs, magnet schools, and charter schools, have become ubiquitous.Choice programs including private school options-including school voucher and tuition tax credit programs-have advanced at a slower pace, but have gained momentum in recent years.
The great hope of market-based reformers is two-fold.First, by creating competition for students, choice reformers hope to provide better schooling opportunities for those choosing to participate.In addition, they hope to create healthy pressure upon the public school system to improve.Scholars have provided evidence that existing choice programs have achieved both of these aims (Greene, 2000;Hoxby, 2001).Second, choice supporters believe that a competitive system for students will lead to the creation of superior school models.Under a near monopoly, no incentive exists to create a better mousetrap.Competition creates this incentive.
Expanded parental choice contains both promise and peril for the Catholic school system.Private choice systems have succeeded in reducing and in some cases fully eliminating the longstanding discrimination in funding against private religious schools.
Preliminary evidence suggests, however, that charter schools are actually threatening to Catholic schools.A RAND Corporation study focusing on the impact of charter schools in Michigan found that private schools were taking a bigger hit from charter school competition than public schools on a student for student basis."Private schools will lose one student for every three students gained in the charter schools" (Toma, Zimmer, & Jones, 2006, pp. 13-14) the study concluded.Ronald Nuzzi, director of the Alliance for Catholic Education Leadership Program at the University of Notre Dame asserted that charter schools "are one of the biggest threats to Catholic schools in the inner city, hands down.How do you compete with an alternative that doesn't cost anything?"(as cited in Meyer, 2007, p. 17).
Ironically, many of the best charter schools drew inspiration from Catholic school practices.A fully scaled system of high performing charter schools for inner-city areas may represent an existential threat to inner-city Catholic schools already struggling with the loss of religious staff and the movement of parishioners to the suburbs.
The Manhattan Institute ranked the 50 states in terms of educational choice, and ranked Arizona at the top (Greene, 2002).Arizona lawmakers have created an extensive system of public school choice, with a strong openenrollment law, public magnet schools, and the nation's strongest charter school law.Arizona lawmakers passed the charter school law in 1994, and over 500 charter schools operate in the state.Arizona also has passed laws to assist with private school choice, including two tax credit programs and two voucher programs.
Arizona's experience with charter schools suggests that charter schools do deliver important educational benefits, including the improvement of public schools, as will be demonstrated below.Furthermore, Arizona's experience provides a counter-example to Michigan in that the Catholic school system has done well despite the proliferation of charter schools.Arizona therefore provides a roadmap as how to expand both public and private choice systems without losing Catholic schools in the process.
The question is not whether lawmakers should pursue either public or private forms of choice-they can and must pursue both simultaneously.

THE CASE FOR SCHOOL CHOICE
Choice advocates, dating back to Milton Friedman, have argued that the traditional public school model is a system of government-run quasi-monopolies.They believe that government has a captive audience in parents who lack the resources to locate in a relatively high-performing school district or pay private school tuition in addition to their public school taxes.Key to improving public school performance, therefore, is to break the government's education monopoly and introduce competition between schools.According to school choice advocates, competition is a powerful incentive for public schools to make better use of resources and to improve academic performance.
High-quality control group studies of the effects of school vouchers show a consistent pattern of small but cumulative academic gains for children participating in choice programs.Other positive effects include much higher parental satisfaction, improved racial integration, increased tolerance, and improved civic values (Greene, 2000).
A fine line exists between stability and stagnation.In education policy, we once were content to have sailed well past that line.Our answer to all education problems was to put in more money.In 1960, the average pupil in American public schools made due with a spending per pupil figure of $375 (around $2,300 in inflation-adjusted dollars; National Center for Education Statistics [NCES], 2005a).Today, they are approaching close to $10,000 in revenue per pupil (U.S. Census Bureau, 2007).Even after adjusting for inflation, spending per pupil in the public school system has more than tripled since the first baby-boomers attended schools.
Our education problems worsened despite the increased spending.Today, 38% of our fourth graders have failed to learn basic reading skills (NCES, 2005b), and around a third of our high school students fail to grad-uate on time with a high school diploma (Orfield, Losen, Wald, & Swanson, 2004).As today's dropouts are largely those students who failed to learn to read in elementary schools, tomorrow's dropouts are already in the pipeline.
Consider this level of stasis in education in comparison to the computer industry.Today, you could literally throw a dart in the computer section of a department store and have it land on a personal computer which is more powerful and cheaper than what was available 2 years ago.By comparison, the public school system continues to plod along, always spending more but often producing less.
The productivity of spending in our public education system has collapsed over the past half century.We spend beyond the dreams of avarice for a public school superintendent of the 1960s, but we do not produce better results.For decades, we have been throwing money at our public schools and failing to notice that students were failing to benefit.Our nation's poorest families cannot afford to buy into high-quality suburban school districts, or to pay private school tuition in addition to their school taxes.Policymakers from both parties have therefore increasingly embraced policies creating options for parents.Nationwide, nearly a fourth of K-12 students will not be attending their neighborhood public schools this fall, opting instead for an array of public and private options-including magnet, charter, private and home schooling.Arizona, Iowa, Ohio, Rhode Island, and Utah have all passed new school choice programs in the last 2 years.For many, especially for inner-city children, however, these options remain far too scarce and this momentum must accelerate.
Charter school operators such as KIPP, Yes Academies, and Amistad have proven definitively that low-income, inner-city children can learn at an accelerated pace, and can even outperform our complacent suburban schools and attend elite universities.These innovators face huge political and practical obstacles in making these schools more widely available.Already, however, they have settled the question of whether we must settle for today's failed status quo: we need not.Our students can learn.We adults simply have to learn how to follow the example of those who are getting the job done.
Our students need a market for K-12 schools.The market mechanism rewards success and either improves or eliminates failure.This has been sorely lacking in the past, and will be increasingly beneficial in the future.The biggest winners will be those suffering most under the status quo.
New technologies and practices, self-paced instruction, and data-based merit pay for instructors may hold enormous promise.Before the current era of choice based reforms, they did not fit the 19th century, unionized model of schooling, so they were not seriously attempted.Bypassing bureaucracy, a new generation has begun to offer their innovative schools directly to parents.Some have already succeeded brilliantly.

DO CHOICE MECHANISMS IMPROVE PUBLIC SCHOOLS?
For many, the desirability of school choice hinges on how choice affects the traditional public school system.Gains among choice participants, after all, could be undone if such a system harms the academic progress of students remaining in traditional public schools.Choice advocates hold that traditional public schools will improve performance in a competitive environment, while opponents believe that public schools will falter under the pressure of having money drained from the public system ("Murray Helps Defeat School Voucher Amendment," 2001).
How school choice affects public schools is one of the most important research questions.The amount of empirical literature on the subject is limited but growing.Evaluations of the privately financed Horizon voucher program, which offered school vouchers to all children in the Edgewood school district in San Antonio, Texas, have found both academic gains and a decided lack of financial pain for the school district (Aguirre & Ladner, 2003;Greene & Forster, 2002).
In the 2001 study "The Rising Tide," Harvard economist Hoxby studied the impact of Arizona charter schools on traditional public schools.Specifically, Hoxby compared the achievement gains in public schools losing 6% or more of their enrollment to charter schools with achievement gains in public schools facing less competition.Hoxby found gains in fourth grade reading four times larger in schools facing competition when compared to those not facing competition.Similarly, academic gains were three times larger in fourth grade mathematics, seven times larger in seventh grade reading, and three times larger in seventh grade mathematics.Hoxby wrote, Let's compare a municipality that did face charter competition, such as Phoenix, with its affluent suburbs.If Phoenix were to maintain its faster rate of improvement, it would close the achievement gap between its students and those in its affluent suburbs in less than ten years.(2001, p. 74) Hoxby found similar results with the Milwaukee Parental Choice Program, and with charter schools in Michigan.In Milwaukee public schools, heavily impacted by voucher competition, reading gains were approximately twice as large in schools impacted by voucher competition as in unaffected schools.Academic gains in science were more than twice as large, and language gains were statistically significant.Similarly, Michigan public schools facing significant charter school competition made fourth grade reading gains twice as large as those not facing competition, with smaller but statistically significant differences found in other subject areas and grades.
GreatSchools.netprovided the school characteristics and academic data from 93 traditional public schools in the Tucson area.The Arizona Department of Education (ADE; 2005) served as the source for enrollment trend data.Using these ADE enrollment figures, schools were divided into those facing significant competition for students and those not.Schools showing a 6% or greater enrollment decline between 2001 and 2004 qualify as facing high levels of competition for students.Twenty-five of the 93 Tucson public schools had an enrollment loss of 6% or more, while the 68 who did not constitute our comparison group.
For purposes of this study, it does not matter whether parents avail themselves of charter schools, other public schools through transfer options, private schools, or home schools.Given Arizona's rapidly expanding student population, it is not only possible but also normal for public schools to show growing enrollments despite the availability of other options.Through open enrollment, schools can gain students to make up for losses.A declining enrollment, in short, is a serious sign of trouble in fast-growing Arizona.Hoxby's research indicates that Phoenix schools responded positively to the challenge.
This analysis replicates the Hoxby research with a different set of schools (Pima County as opposed to Phoenix) and from a more recent period.Figure 1 illustrates the average national percentile ranking gains made by these two groups of Tucson schools during the 2001-2004 period.In Stanford 9 reading scores, Tucson-area public schools facing competition gained an average four national percentile points, while the comparison group's gains averaged less than one national percentile.Overall, the academic gains of the competition group of schools were approximately 5.4 times larger than the comparison group.In mathematics, Tucson public schools facing competition for students made Stanford 9 gains approximately twice as large as those of the comparison group did.
The competition group's gains on the Stanford 9 language arts exam are more than 13 times greater than the comparison group's gains, as illustrated in Figure 1.The academic gains of schools facing competition for students are similar to those found by Hoxby (2001).These differences in scores look impressive, but we must subject them to a multivariate analysis before drawing conclusions.A regression analysis will establish whether these differences are of sufficient size to be statistically significant.Second, the regression technique can statistically control for a variety of other factors that may explain the differences between these two sets of schools.
For example, it is possible that the schools facing greater competition also have a smaller percentage of economically disadvantaged students than the comparison group schools.It is also possible that they otherwise vary in the demographic profile of students or teachers.Perhaps these schools had more experienced teaching staffs or lower teacher-student ratios, for example.Fortunately, the regression technique allows us to separately control for each of these factors, while still including an independent variable for competition, again measured as the decline in enrollment.
Table 1 presents the results for three regression models, using the respective gains on Stanford 9 reading, mathematics, and language arts score gains as the dependent variables, respectively.The competition variable displays a consistent statistically significant relationship with national percentile ranking gains in all three academic areas.Meanwhile, the various student demographic variables (poverty, ethnicity) and teacher characteristic variables (teacher experience and education) and the student/teacher ratio variable all fail to demonstrate a significant relationship with academic gains.The only other significant result is for the percentage of teachers with 7 or more years' experience on mathematics, and it is a negative result, meaning that schools with more experienced teachers had lower gains on mathematics scores.A similar analysis (Ladner, 2007) of 408 Phoenix-area elementary schools also found significantly higher academic gains on Stanford 9 language arts, mathematics, and reading exams during the 2001-2004 period.Of the 408 number of schools in the Phoenix sample, only 9 elementary schools endured a loss of more than 6% of students between 2001 and 2004.While the Phoenix schools facing this level of competition made significantly greater gains on the Stanford 9 language arts, mathematics and reading scores in comparison to the other schools, the greater percentage of Tucson schools facing such competition made for a more robust test of the competition hypothesis.Collectively, these results strongly reinforce the findings of Hoxby's (2001) research in showing that competition for students creates positive pressure on schools to improve academic performance.When faced with such competition, schools in both Tucson and Phoenix improved their academic outcomes at a significantly faster rate than schools not facing competition.

HAS PUBLIC SCHOOL CHOICE HURT ARIZONA CATHOLIC SCHOOLS?
Fortunately, in addition to public school choice, Arizona lawmakers have created private school choice programs.Three years after the passage of the charter school law, Arizona Governor Fyfe Symington signed the nation's first scholarship tax credit law.This law provided a dollar for dollar tax credit to individuals or couples donating to a private scholarship group, known as a scholarship tuition organization (STO).The maximum size of the credit is $500 ($1,000 for a couple filing jointly).
Last year, Arizona STOs raised $51 million for scholarships.STOs affiliated with Catholic schools raised over $15.5 million and some unknown number of students attended Catholic schools with scholarships from other STOs (Arizona Department of Revenue, 2007).In 2005, Arizona lawmakers passed a law known as the Marriage Penalty Elimination Act.This law's impact on the individual scholarship tax credit was to phase in an increase in the maximum donation a couple could make from $625 to $1,000 in 2006 and beyond.
The tax credit program seems to have aided in Arizona's Catholic schools resisting the national trend toward declining enrollments despite the proliferation of charter schools.For example, during the 2004-2006 period, schools in the Diocese of Phoenix saw a 2% increase in enrollment against a national decline.Two new Catholic schools opened in the 2006-2007 school year, with four more scheduled to open in the near future (Hammel, 2006).
Better still, in the summer of the 2006, The Catholic Sun reported that in the Diocese of Phoenix "an increasing demand for Catholic education coupled with aging campuses has meant that every Catholic high school in the diocese has either recently completed or is in the midst of a capital campaign" (Junker, 2006, ¶3).This is a good problem to have.
In addition, in 2006 Arizona lawmakers passed a corporate scholarship tax credit to supplement the individual credit.The corporate credit has a beginning total cap of $10 million which expands by 20% each year until it reaches over $20 million.Also in 2006, Arizona lawmakers passed two limited school voucher programs aimed at two very disadvantaged student groups-students with disabilities and students who have been in the foster care system.
Arizona has therefore struck a more appealing balance between private and public school choice than Michigan.Sadly, Michigan's constitution essentially forbids private school choice of any sort, and the Archdiocese of Detroit has witnessed a 20% decline in enrollment since 2002 and currently faces another round of school closures.Overall, 29 schools in the archdiocese have already closed (Bouffard, 2007).
Choice programs do not represent the only difference between Arizona and Michigan of course.Arizona has experienced rapid population growth in recent years, while Michigan's school-age population has remained essentially flat.As of November 2005, however, Arizona had 449 charter schools operating, while Michigan had only 239 (Center for Education Reform, 2006).Total charter school enrollment is 12.5% higher in Arizona than in Michigan, despite the fact that Michigan's population is 70% larger than Arizona's.
In short, if charter schools are having an adverse impact on Catholic school enrollment in Michigan, they ought to be having a similar impact on schools in Arizona, with population growth and private choice programs serving as mitigating factors.Fortunately, this experience suggests that it is possible to have a vigorous charter school law, with all of the benefits discussed above, along with a thriving Catholic school system.States with slower population growth than Arizona, however, should do more on the private school side than Arizona has done to date.

FUNDING NEUTRALITY BETWEEN SCHOOL SECTORS
While Arizona ranks first in school choice, we should fully recognize that even Arizona has taken only incremental steps toward full parental choice.As demonstrated above, those steps have created positive results in the public school system.
STOs granted scholarships with an average amount of $1,643 per pupil in 2006, although that amount should improve given recent changes in the tax credit laws.By comparison, Arizona taxpayers grant almost $7,000 per pupil for students in charter schools, and over $8,000 in revenue per pupil for students attending traditional public schools (Arizona Department of Revenue, 2007).
Arizona's current education policy seems to suggest that a student attending a public school serves the public interest in a fashion more than five times greater than a child attending a private school.This is implausible, to say the least, especially given a large amount of high quality research suggesting positive effects of participating in choice programs (Greene, 2000).These benefits not only include academic gains, but also much higher rates of parental satisfaction, better racial integration, and higher civic values.
Private schools serve the public interest, and on average do so quite well.State education policy should therefore be entirely neutral as to whether a parent chooses a public or private school for his or her child.
A universal tax credit represents a far more powerful vehicle for promoting school choice.Some school choice advocates believe that a universal tax credit represents the best school choice alternative (Anderson, McLellan, Overton, & Wolfram, 1997).A universal credit essentially combines a scholarship credit with a personal use credit.Under a universal credit, taxpayers can claim a credit for a donation to a scholarship tuition organization (a scholarship credit) and for private school expenses associated with their own child (a personal use credit).Thus far, we lack an operating example of a universal education tax credit, but there is little doubt that it can be a powerful mechanism for increasing parental choice in education.
Other choice proponents prefer a school voucher mechanism to level the funding playing field between public and private schools.Vouchers have operated successfully in Milwaukee since 1990.Lawmakers in Arizona, Florida, Ohio, and Utah passed vouchers in recent years, and Congress passed a program for Washington, DC.Vouchers represent a more straightforward method to achieve choice than tax credits, and are generally likely to be easier for parents to use.Some state constitutions express greater hostility toward vouchers than tax credits, however.Such mechanisms are not mutually exclusive, as both Arizona and Florida have passed both voucher and tax credit laws.Education reformers should pursue both reforms-in fact it may be possible to combine them.
Parents must pay public school taxes even if they do their fellow taxpayers the service of placing their children in a private school at their own expense.If parents decide to seek an education they find more suitable for their children, they effectively pay twice-once when they pay taxes, another when they pay tuition and fees.A personal use credit can reduce this double payment penalty, expanding access to private schooling.In the process, such a credit could improve the performance of public schools by expanding competition for students and reducing public school overcrowding.
Personal use credits do face limitations.First, policymakers would find it challenging to develop a personal use credit that both made a meaningful impact and provided greater aid to lower-income families.Lower-income families, for example, have smaller income tax liabilities and thus a smaller potential to benefit from a personal use credit.
A number of possible strategies could mitigate this problem.First, lawmakers could design the personal use credit to be refundable.Refundable credits return money to the taxpayer even if the amount of the credit exceeds liability.Second, lawmakers could create a separate voucher program to provide additional aid to low-income families.A dual system of a personal use credit accompanied by a means-tested voucher would provide a system of universal choice while clearly advantaging economically disadvantaged families, who would be the only people eligible to benefit from both the personal use and voucher programs.A third possibility would involve enhancing both the individual and corporate scholarship tax credits and applying a means test to the eligibility of the scholarship tax credit in concert with the creation of a personal use credit.
The RAND Corporation report on Michigan (Toma et al., 2006) provides an important cautionary tale regarding the pursuit of a purely public system of school choice.Michigan's circumstances, however, are quite unusual.Michigan has a stagnant school-age population, severe levels of racial segregation in the Detroit area, a strong charter school law, and a constitutional ban on any sort of choice program including private schools.
Arizona's circumstances are unusual as well-the most vigorous charter school law, a rapidly expanding school age population, and some limited private choice programs.Under these circumstances, Arizona Catholic schools have more than held their own.The experience of these two states-both enjoying the large benefits of charter schools, but with starkly different trends in Catholic schooling-suggest strongly that choice supporters must continue to seek both charter school and private choice laws, but redouble their efforts on the private choice programs.

Figure 1 .
Figure 1.Relative 5th Grade Stanford 9 National Percentile Ranking Gains for Pima County Public Schools

Table 1
Impact of Competition on Stanford 9 Reading, Mathematics and Language ArtsGains,  2001Gains,   -2004 Note.Ordinary Least Squares regression; entries are unstandardized coefficients; standard errors are in parentheses.*p <0.05 **p <0.01 Data sources: GreatSchools.net,Arizona Department of Education.