Manhattan Institute for Policy Research.
Subscribe   Subscribe   MI on Facebook Find us on Twitter Find us on Instagram      

Civic Report
No. 10 June 2000

Does Charter School Competition Improve Traditional Public Schools?

Paul Teske
Professor and Director of Graduate Studies, Department of Political Science, SUNY-Stony Brook

Mark Schneider
Professor and Chair, Department of Political Science, SUNY-Stony Brook

Jack Buckley
PhD candidate, Department of Political Science, SUNY-Stony Brook

Sara Clark
PhD candidate, Department of Political Science, SUNY-Stony Brook


Of the many school choice initiatives spreading throughout the country, charter schools are by far the most common. As of January 2000, over 1,700 charter schools serving more than 350,000 students have been created in nearly 40 states. That these schools are popular among parents can be seen by a simple fact: approximately 70 percent report having waiting lists. However, it’s critical to remember that despite the rapid growth of charter schools, nearly 90 percent of American children continue to be enrolled in “traditional” public schools.1  Since this distribution of students across sectors is likely to change only modestly in the foreseeable future, we believe that one of the most important issues of the many raised by the charter school explosion is the extent to which charter schools affect the behavior of traditional public schools by competing with them for students.


We studied Springfield and Worcester, Massachusetts, Jersey City and Trenton, New Jersey, and the District of Columbia (D.C.), to appraise how charter schools are affecting the traditional public school system.2 We found that charter competition has not induced large changes in district-wide operations, despite the fact that a significant number of students have left district schools for charter schools.

This may be explained by the fact that state policies generally cushion districts from the financial effects of departing students. Demographics play a role, too. Rising total enrollments, a by-product of the “baby boom echo,” and recent high rates of immigration, have helped districts avoid fiscal pain; some have maintained absolute enrollment (and budget) levels even as their market share has shrunk.

Districts also cushion individual schools from the financial impact of declining enrollments and shrinking market share. They tend to provide constant resources to shrinking schools, in some cases sending extra money to failing schools to “prop them up.”

Even though state and district policies shielded public school systems from the effects of charter school competition, impeding a full test of the proposition that competition leads to better schools, we found that many superintendents and principals are responding even to muffled competition by making changes designed to produce more appealing and effective schools. This is especially true in districts where the superintendent was already disposed to reform district operations. There the likelihood of such change was enhanced by competition from charter schools. For example, Springfield Superintendent Peter Negroni has replaced more than a dozen of the district’s 42 principals and emphasized school-level accountability for pupil performance. Trenton Superintendent James Lytle has developed a school-based budget process that will allocate funds to schools on a per-pupil basis, ultimately translating enrollment losses into financial losses.

Superintendents have also made changes in response to specific features of charter schools that are attractive to parents. Examples include:

• Springfield is starting a Montessori-style school and one that makes extensive use of laptop computers, both ideas already implemented in popular local charter schools;
• The widespread belief among parents that the Sabis charter school is improving children’s test scores led several Springfield schools to start Saturday study programs, mimicking those at Sabis;
• Two Trenton elementary schools located in neighborhoods most affected by new charter schools are changing from K-6 to K-8 because of parents’ concerns about safety in the public middle schools; and
• In response to widespread parental demand for charter-style before- and after-school programs, Worcester Superintendent James Caradonio and his principals are working with the teachers’ union to replicate such programs in district schools.

Furthermore, we found evidence that school managers respond to competition from other schools. In surveying principals about whether they were changing their school operations in response to competition, we learned that:

• Principals adopt more innovations at their school in direct proportion to the competitive enrollment pressure that they feel;
• As that pressure mounts, principals try hard to boost school efficiency;
• As competitive pressure builds, principals are likelier to feel they do not have enough autonomy to run their schools as they judge best.

These building-level patterns are evident even in districts where the superintendent is hostile to charter schools, such as Worcester.

In the District of Columbia, we learned that parents believe that charter schools have superior physical facilities than district public schools. Washington parents are also more likely to say that charter schools feel safe and that their staff is friendlier and more helpful. These factors probably help to explain why nine percent of all D.C. schoolchildren are currently enrolled in charter schools, one of the highest proportions in the country.

The fact that traditional schools attempt to limit the effects of competition and respond to competition in a piecemeal, rather than a wholesale, fashion may disappoint charter school advocates. To understand why response occurs in this way, we first need to address the theories that underlie the competition hypothesis and show how our observations diverge from these expectations.


From our review of the literature on how competition can affect the behavior of organizations, we identify two theories explaining how the presence of charter schools can leverage change in the public schools. The first theory is based on the diffusion of innovation; the second is based on the indirect financial effect of losing students.

How New Ideas Can Leverage Change

Many proponents of charter schools argue that, because of their greater freedom and fewer bureaucratic rules, charter schools can be “laboratories” for change and experimentation that will provide examples for the reform of the traditional public schools. Many charter school advocates believe that, given this freedom, charter schools will design new and original curricula and programs, that they will experiment with new models of school organization, and that they will develop new methods to encourage parental involvement. Of these experiments, those proving successful will be conveyed to the education community at large and eventually adopted by the traditional public schools (Kolderie 1990; Nathan 1996).

However, at least one assumption in this line of argument may not hold true. There is an expectation that the lines of communication between the two sectors will be open and that information will flow freely between them. As we demonstrate below, the attitude of school district officials towards charter schools varies widely; districts that are hostile to charter schools are unlikely to encourage communication. There are also concerns about whether information flows in the other direction. An official of the District of Columbia Public Schools (DCPS) suggested that there is little actual communication between sectors because there is no incentive for educators at charter schools to convey information back to the traditional public schools, as they are too busy and because many of them have little desire to communicate.

How Financial Pressure Can Leverage Change

The second theory relates to the direct effects of market competition. Specifically, we might expect that the traditional public schools will respond as they lose students to the charter schools if there are financial penalties for this loss, as there are for firms that lose customers or market share. In this view, the elements necessary to propel change are competition and the prospect of financial reward or penalty based on performance and “customer” behavior. This financial incentive could aid in the flow of information concerning successful education programs and techniques between charter and traditional schools: If traditional schools face a loss of budget resulting from declining market share, the incentive to adopt programs that have been proven successful elsewhere should increase.

Indeed, charter school laws around the country have effectively “codified” this argument by mandating that some percentage of per pupil educational expenses follow the child to the charter school, leaving less money in the traditional public school district (though also leaving fewer children to educate). Many advocates have assumed that this financial loss would motivate the traditional public schools to improve performance. As we show below, in reality, such losses have not yet been felt in many districts—mitigating the strength of the “market share” argument.

Some skeptical analysts have questioned whether insulated urban school districts, often run by bureaucrats, really can be changed by these incentives, especially when other techniques tried in the past did not succeed (e.g. Rothstein 1998). The broader literature on the effect of competition on public organizations shows that responses can be more complex than the above arguments indicate. Responses from the traditional public sector can range from outright hostility and resistance, to various degrees of realistic adaptation, including the use of competition to achieve other organizational goals, through attempts to co-opt the competitive process to the advantage of the organization or its sponsors.3  Indeed, we find that the school districts we studied have responded to the increased competition from charter schools in ways that span the spectrum cited in the literature.


To gauge the influence of charter schools and district-level responsiveness, we interviewed district superintendents, school board leaders, charter school heads, teachers’ union leaders, and other district leaders (such as parental placement center heads and school principals). We searched various newspapers and databases to get “third-party” information about local politics, population trends, and issues surrounding the public and charter schools in these districts. We also examined the written materials and websites prepared by state education departments and the school districts themselves. We combined these interview and other data with district-level objective data on enrollment trends, demographics, and test scores to get a more complete picture of each district.

In addition to this district-level information, we surveyed school principals by mail, e-mail, and telephone. The response we received to the survey of principals depended greatly on the degree of support we received from the district superintendent in carrying out the survey, which varied across districts. At one extreme, cooperation in the Worcester and Trenton school districts was superb. In both cities, the superintendents took an active role in distributing our questionnaires (both in printed and electronic versions) and asked their principals several times to participate in the study. At the other extreme, our efforts to survey principals in Washington D.C. were a dismal failure. We tried both regular mail and e-mail approaches to principals in D.C., but to no avail. Eventually, we were told that officials at the central office had instructed principals not to respond. As evident in Table 1, the support of the superintendents clearly affected how many principals responded to our survey.4 

Given the low response rate in Washington D.C., we do not include these principals in some of the analyses presented below.5 

However, given the large number of charter schools in the District, we were able to address two other questions that could not be studied in the other districts. These include a more detailed comparison of the “consumer” friendliness of both types of public school and an analysis of the “package” of reforms evident in the charter schools. With this background in place, we present our most important findings.


Finding: District-Level Attitudes Toward Charter Schools Vary Widely

Although our districts were not chosen as a random sample, nor were they selected to increase the variance on the dependent variable, we found a wide range of district-level response to charter schools. Figure 1 illustrates how support for charter schools varies across the five cities. Starting with the strongest level of support in Springfield, we give some illustrations of how district-level support varies across our five cities.

In Springfield, support for charter schools is extensive among district officials. Indeed, Superintendent Negroni and a member of the elected city school board actually sit on the Board of Directors of the Sabis International Charter School, the largest charter school in Springfield. The support for the Sabis Charter School is also evident among the city’s political elite: the Mayor of Springfield serves as a director of the Sabis school as well.

In our interviews with school district officials we found virtually no hostility directed at charter schools. To the contrary: Negroni considers children who attend charter schools to be “our”6  children, since they live in Springfield. He views charter schools as part of a “toolbox” of techniques available to him to leverage reform in his district and he argues that “beating Sabis” has now become a mantra in the district, energizing the public schools to perform better.

In Trenton, while no school official serves on any charter school board, district leaders support charter schools as a form of competition they can use to leverage further reform activity by their own principals. Lytle notes, “I believe in competition. We intend to be in the forefront of urban education reform. We welcome this competition and believe we’ll all be better for this.”(Shoemaker 1998).

While Springfield and Trenton represent by far the most positive responses to charter schools among our five sites, Jersey City demonstrates a fairly “neutral” response to charter school competition. As noted in Appendix 3, the Jersey City superintendent, Richard DiPatri, was appointed by the state of New Jersey, which took over the city schools ten years ago. Officials who we interviewed in the Jersey City public school district cooperate with charter schools but do not see them as strongly positive or negative influences on their actions, and they continue to implement their own reforms, without seeing charter schools as a leverage point to further their own agendas. According to DiPatri: “We cooperate with the charter schools, but our business, and our state mandate, is to improve the whole Jersey City public school system.”

Leaders in the two other districts, Worcester and Washington D.C., are negative towards charters and have actively worked against them.

Worcester officials, including the superintendent, are actively opposed to charters and are leading a lobby of municipalities in Massachusetts to get state rules governing charters changed. There are several main issues that motivate their opposition, among the most basic is the claim that parental choice of charter schools is motivated by racial concerns rather than by academic or programmatic considerations. Indeed, one official went so far as to argue that “choice is race.”

In addition, most of the Worcester officials we interviewed argued that the charter schools do not “add value” to the education of their students and that the charters represented an unnecessary drain on the resources of a district that they argue is actually performing quite well. Moreover, echoing a common charge against charter schools, several officials we interviewed argued that the charter schools, despite laws to the contrary, were screening students and either not admitting or “counseling out” high cost students, such as students in need of special education.

In D.C., although official statements from the office of Superintendent Arlene Ackerman reflect a neutral attitude toward the charter schools, in fact, one of the primary ways in which the DCPS demonstrates opposition is by increasing the level of difficulty charter school organizers face in locating suitable facilities. Securing an acceptable facility is one of the most pressing problems facing new charter schools throughout the country, and despite the Congressional mandate to give District charter schools priority in obtaining vacant public school buildings, new charters continue to face an alarming level of difficulty in exercising this priority. In their 1999 report, Henig et al. cite several examples of “puzzling” occurrences surrounding the sudden unavailability of previously available and occasionally already-promised school facilities. For example, the Hyde Park charter school negotiated for 15 months with D.C. school officials to lease a vacant school building, with an option to buy it in 10 years. However, when the plan was submitted to the Board of Trustees for approval, the lease process was halted without explanation. Hyde Park school officials were eventually informed that the lease price would increase and that there was no guarantee of a building purchase. Council member Kevin Chavous told the Washington Post that: “It sounds like [school officials] reneged on a deal.” (Wilgoren 1998).

Another problem has been obtaining charters from the D.C. Board of Education. For example, the Kwame Nkrumah charter school was told that the Board had approved a charter, and they prepared to open. But, just days before the school year, the Board voted against a charter in a closed meeting, and sent security officials to tell students that their school was not authorized (Strauss 1999).

D.C. charter school leaders believe that city officials are slow and erratic in their responses because they are worried that students and parents will favor the charters and leave the traditional system in even greater numbers. Superintendent Ackerman and other district leaders have expressed concerns about loss of financing to charter school competitors, and argue that such competition is not fair.

Finding: District Level Support is Not Related to Market Share Held by Charter Schools

As noted, a market share approach to charter schools would link the percentage of students attending charter schools to the behavior of officials in the traditional public schools. We hypothesized that as the market share of charters increased, the resulting loss of market share for traditional schools would lead district school officials to take an antagonistic view of their competition.

However, as Figure 2 shows, the attitude of school district leadership to charter schools is not a direct function of the loss of market share. We believe that there are two reasons for this lack of relationship:

•First, districts and their individual schools have to date been shielded from most of the costs of losing students to charter schools. In short, despite the compelling logic linking market share, competition and financial benefits or losses to organizational responsiveness, school districts have to date been essentially held harmless fiscally from the growth of charter schools.
•Second, the attitude of the district superintendent and, through the superintendent, the attitudes of other high level administrators seem to be more a function of their individual beliefs rather than a function of the current level of charter school enrollment in their district.

We believe that both of these factors have reduced the effects of competition from charter schools and changing “market share.”

Finding: School Districts Have Been Shielded from the Fiscal Implications of Losing Students

Theoretically, the loss of students in the traditional public schools to charter schools would be accompanied by a corresponding loss of resources. When districts start to lose a significant share of students, they are likely to feel the financial shortfalls. This would then lead to the restructuring of failing public schools to attract students back and, in the extreme, a district-level decision to close (the equivalent of “bankruptcy”) those schools that cannot or do not respond.

In our fieldwork, however, we found that in every district we studied a variety of financial and policy mechanisms have shielded the districts from most financial losses and, at the state or district level, compensatory financial mechanisms have been put in place to soften the impact of the growth of charters.

For example, in Massachusetts, we found almost universal agreement among district leaders and principals that the financial impacts of growing charter school enrollment have been negligible. This is partly because a Massachusetts law enacted in 1995 and amended in 1999 created a sliding scale of state reimbursement for financial losses due to charter school enrollment.7  The amount reimbursed decreases over four years after the new school opens its doors (from 100 percent to 60 percent to 40 percent to 0 percent). This slow adoption of full financial penalty for loss of students, along with additional state aid to the schools through other channels, has thus far prevented the traditional schools in Massachusetts from experiencing the true effects of competition from the charter schools.8 

The schools in the two New Jersey districts have also avoided the full financial losses that should flow from growing charter enrollments. As a consequence of Abbott v. Burke,9  the latest N.J. Supreme Court decision in a 25-year long educational funding legal saga, the 30 poorest districts in New Jersey, which include Trenton and Jersey City, are entitled to a large share of financial resources from the state. These extra funds support new pre-K programs, reduce class size, and fund a variety of other compensatory programs. To make sure that these funds are well spent, the court has mandated a set of reforms known as “Whole School Reform” (WSR). Thus, Trenton and Jersey City are receiving new funds to implement WSR far in excess of the amounts they have lost to the charter schools. Most notably, Trenton received more than $10 million for WSR, and by becoming a “lighthouse” district for WSR implementation, is attracting consulting funds and expects to gain even more discretionary state monies.

Additionally, the state of New Jersey is expected to provide substantial funding for new school construction and school modernization, which are critical issues in Trenton and Jersey City. As a result, even with the outflow of students to charter schools, the schools have not experienced any severe financial repercussions.

In D.C., budgets of both traditional and charter schools are supposed to be determined by a uniform per-student formula that includes a base allotment of $5,500 per pupil, with extra monies available for various add-ons and facilities allowances. However, federal legislation permits block appropriation of funds to the public schools for “special needs,” thereby allowing individual school budgets to be enlarged or held constant despite drops in enrollment, and, at least so far, preventing the district from experiencing an overall budget decrease. Henig et al. observe that “With enrollment levels in public charter schools expected to grow, it is unclear how schools will be able to meet their financial responsibilities without sacrificing their educational missions. Luckily for both DCPS and public charter schools, the size of the elementary and secondary education budget has grown substantially in recent years, easing, at least for the time being, the potential for intense competition between the two types of schools.” (1999, 38). Again, this may be changing: As the Washington Post recently reported, “[l]ately, city officials have been warning [Ackerman] that the school system could wind up with millions of dollars less than expected for the 1999-2000 school year, in part because of the growing popularity of new D.C. public charter schools—which are drawing public finds away from traditional schools.”(Strauss 2000)10  Thus while in the future the financial costs of losing students to charter schools may force a response, to date, these costs have not yet been fully imposed.

Finding: To Date Individual Schools Have Also Been Held Harmless

Just as no district has yet been held fully accountable for losses of students to charter schools, we find that no district has yet held individual schools fully accountable for their losses of students to charter schools. For most districts, in the long run money will have to have some reasonable relationship to enrollment in particular schools. Indeed, in some of these districts, at least a portion of school-level budgets is tied to the number of pupils served—but probably not yet enough to force reforms. And this budgeting trend often comes into conflict with a much more traditional response: Superintendents often redistribute money to failing schools to try to “prop them up.” In some districts the resolution of these two forces may be changing in favor of holding individual schools more responsible.

For example, only a few principals in Trenton reported feeling budget constraints this year associated with loss of students, but many more expected to feel a large budget pinch next year as a result of choice. This is not surprising, because Superintendent Lytle plans to implement a flexible (50-60 percent “cash-based”) school-based budget process that will allocate funding on a per pupil basis, with decisions made by a team of administrators, teachers and parents. With this kind of budget process in place, Trenton will translate school-level losses of students into financial cuts.

In the other districts, however, we did not hear much about any efforts to apportion cuts to specific schools. This is not because districts are ignoring school-level activities. In fact, in several of these districts, superintendents are holding principals more accountable. In Springfield, Negroni has replaced more than a dozen of his 42 principals and emphasizes the “District 2” model of choice and school-level responsibility that he learned from Anthony Alvarado in New York City before moving to Springfield. Caradonio in Worcester has developed an elaborate school-level accountability tracking system; he has not replaced large numbers of principals, because he believes that he and his predecessor put a strong cadre in place. DiPartri made principal accountability a centerpiece of his reform in Jersey City, appointing a new supervisor of principals. But in general, while district superintendents must ultimately care about how individual schools are being run, so far, they are not explicitly punishing schools for losing students to charter schools.

Finding: Population Trends May Also Be Blunting the Impact of Charter Schools

Despite the general decline in inner-city populations over the past few decades, all of our districts, as a result of both the “echo baby boom” and recent immigration, have witnessed increases in total enrollment (See Appendix 1). This has acted to relieve the traditional schools of some of the competitive threat from charter schools as new students quickly fill seats vacated by those opting for a charter school. Thus, even if enrollment in charter schools increases, the traditional public schools do not necessarily experience a net decline in the number of students in attendance. Even in the District of Columbia, for example, the total number of students in all public schools has increased over the last decade. As Figure 3 illustrates, the growing number of charter school students is taking an increasing proportion of total enrollment, but even then, total public school enrollment was higher at the end of the decade than it was at the beginning.

While D.C. does have surplus buildings, none of the New Jersey or Massachusetts districts have excess space to accommodate growing enrollments. Trenton, for example, is already sending its special education children to other districts because it lacks space to house them. In both Jersey City and Springfield, school district leaders explicitly regarded charter schools as a way of avoiding large capital outlays to house an expanding student population.

In short, with many urban districts currently experiencing growth, charter schools can be viewed as a positive way to deal with space problems. The fact that many schools are not seeing fewer students, even as charter schools are created, blunts the competitive pressure they feel from expanding charter schools.


Finding: The Role of School Leadership is Critical in Reform

The “effective schools” literature and many other studies of high-performing schools argue that school leadership is critical to performance (e.g., Teske and Schneider 1999; Chubb and Moe 1990; Purkey and Smith 1983, Byrk, Lee and Holland 1993). According to Hess (1999b, 7): “School improvement requires time, focus, and the commitment of core personnel. To succeed, the leadership must focus on selected reforms and then nurture those efforts in the schools.” Thus, we might expect to see variation in the response of school districts based on differences in leadership styles.

In our fieldwork, we found that district leaders and principals who are entrepreneurial and reform-oriented are using charter schools as a tool to increase their leverage over their schools and force them to institute new programs and improve performance. We also found that the predisposition of superintendents toward competition seems to precede the growth of charters.11

Of all the district leaders we interviewed, Springfield Superintendent Negroni was the most vocal supporter of charter schools. Given his background, his support for choice is not surprising. For example, Negroni worked with Anthony Alvarado, one of New York City’s most successful advocates of choice, and was hired by Springfield a decade ago with the explicit mandate to reform the Springfield public schools. Indeed, he talks about training his principals in the expanded choice models that Alvarado pioneered in New York City’s Districts 2 and 4. Therefore it is not surprising that Negroni regards charters as “a spark plug to accelerate change in the system,” and that he argues charter schools “get people’s attention” and provide additional leverage for reform. And it is not surprising that he has borrowed ideas developed in local charter schools: Springfield’s traditional public schools will soon include a Montessori-type school and a school based around laptop computers, a key element of the Edison Project’s approach. Perhaps most importantly, Negroni believes that his principals have the power and the flexibility necessary to improve the quality of services their schools provide—and that charters are forcing them to use that authority to make beneficial changes.

Similarly, in Trenton, Lytle had an explicit mandate to reform the public schools and, as noted earlier, had in fact written at least one application to create a charter school. Lytle appreciates the increasingly competitive environment in which public schools are operating. In his first speech to Trenton principals he pointed out that he could, in effect, “sell” the whole system to the Edison Project, who could run it cheaper and perhaps better than it was being run. Lytle was making an explicit market competition argument to support his push for reform.

In contrast, Worcester Superintendent Caradonio took over a relatively stable school district after serving for over 10 years as the Deputy Superintendent to James Garvey, a popular and well-regarded superintendent. He simply does not see the need for wholesale reform, and views the resources and attention given to charter schools as harming his progress towards incremental improvements.

Although DiPatri had been involved with charter school legislation as a state official, he felt that his state mandate in Jersey City was to clean up the previous mess, and to do so with a centralizing approach, emphasizing top-down accountability. For example, he has implemented a principal mentoring program and an academy for the preparation of principals. He appointed five new principals in the latest school year. While not at all hostile to charters, he views them as irrelevant, at best, to his reform agenda.

Finally, in D.C., Superintendent Ackerman has been working hard to reform, and to improve the image of, the traditional public schools. But, she has been criticized for putting more resources into attempting to improve the worst schools, at the expense of maintaining the good ones. And, some principals feel that she has increased their bureaucratic burdens, while not forming strong and useful relationships with them.12

Finding: Even if Districts Remain Insulated, Principals Respond to Pressure

While superintendents clearly set the tone for their school districts and set the general course of reform for the individual schools in their district, we found evidence from our survey of principals that the sound of parent footsteps heading to charter schools is being heard at the school level, regardless of what their superintendent says about charters. As front-line managers, many principals recognize that the funding of their schools and even their job tenure will increasingly be tied to enrollment trends, so perhaps this is not surprising—and may be occurring even as superintendents say that a response is not necessary.

We found several patterns in our surveys that suggest that the growth in charter school enrollments is affecting how principals behave. Here we report data from our survey of 76 principals in the four districts with reasonable response rates (that is, we exclude D.C.).

As Figure 4 illustrates, there is a relationship between the principal’s expectation of loss of students to charter schools and other forms of competition in the near future and the number of recent reforms introduced or expanded in a school.13  Similarly, as expected losses to competitors increase, principals report spending more of their time on improving the efficiency of their schools (see Figure 5). And Figure 6 shows that principals who fear the loss of students to competitors are also more likely to feel that they do not have enough autonomy to implement the kinds of changes and policies that they see as necessary to make their schools perform better and more competitively.

In short, principals feeling competition are trying to act more entrepreneurially: enhancing efficiency, changing the programmatic mix of their schools and expressing frustration over restrictions on their ability to do more. It is important to note that these same patterns generated in the aggregate data also appear among principals in a district like Worcester, where the superintendent and other district-level officials downplay the challenges represented by growing charter school enrollments. So, principals, the front-line managers, may be feeling the effects more strongly, and in turn may be trying to react more quickly, than are district officials.

The pressure on public school principals is likely to increase, especially in districts where the superintendent is intent on reform. For example, in Massachusetts, principals are no longer tenured and now serve on contracts that last from one to three years, due to provisions in the state’s 1993 Education Reform Act. Since the law was enacted, in Springfield, Superintendent Negroni has used this new power to a considerable degree: he reports that he forced out principals from about a dozen (of 42) schools since the 1993 enactment.


The literature supporting charter schools often portrays them as venues in which new teaching techniques and new packages of services will be pioneered and then diffused to the traditional public schools. In the most extreme forms, many advocates of charter schools view them as “shining lights” that will illuminate the future of education. Indeed, for many proponents of charters, this diffusion of ideas through example is as important as any theory of market competition.

Contrary to this view, we found a sharp divide between leaders in the traditional public schools and the charter schools that has so far muted any diffusion effects of charter school reforms. At the most fundamental level, many leaders in the traditional public schools feel that the environment in which charter schools operate is not the same one they face. And many public school leaders think that the charter schools are not doing anything new or innovative that they wish to emulate.

Finding: There is Hostility Between the Sectors that Limits the Spillover from Charter Schools to Traditional Public Schools

Significantly, we found that there is often a fundamental hostility between the traditional public schools and the charter schools. Even in districts where the level of hostility is low we found little evidence that schools in either sector have reached out to schools in the other sector. Consequently there is little evidence of cooperation and little evidence of shared ideas and information in any of the districts we studied. In many ways, each side feels that the other side has great advantages that make for an “uneven playing field” and limit the lessons that can be learned across sectors.

Not surprisingly, district officials in Worcester were quite aware of this gap, but even with the tight cross-sector links, they still note that there is no information shared between the charters and the district. Indeed, there does not even seem to be much interest in what the charters are doing.

Worcester’s Superintendent Caradonio does not believe that charters are real laboratory schools introducing real innovations, and consequently does not spend any time learning exactly what they are doing differently. In addition, district and union leaders share the view that charters are “union busters.” Mark Brophy, the president of the Worcester teachers’ union, called them, “a conspiracy to implode public education.” Springfield teachers’ union President Tim Collins, echoed these sentiments, adding that he believes charters create a two-tier system that exacerbates differences based on race and socio-economic status.

The Rosenblum Brigham report (1998) on Massachusetts charter schools shows that the attitudes found in Worcester are not unique. It found that: “There is no significant sharing or dissemination of practices from charter schools to district schools at this time. The reasons range from the practical (a lack of mechanisms for sharing) to the philosophical (there is a sense of competition and even hostility toward the charter school system that precludes sharing).”

In D.C., a high ranking official in the DCPS echoed this, arguing that “there is no incentive for charters to convey information back to traditional public schools about what is and isn’t working for them. Administrators, faculty, and staff at charters either have too much to do to have time to do this, or they have consciously left the traditional system so they don’t want to communicate with DCPS regarding successful and/or failed programs.”

One fundamental source of the suspicion that leaders of traditional public schools show for charter schools is that, despite the laws and empirical evidence to the contrary, many public school officials believe that charter schools educate a different population than the traditional schools do. For example, echoing a common charge, officials in Jersey City argued that charter schools can effectively force or “counsel” out problem students, who them return to the responsibility of the traditional public schools. Similarly, in a front-page Education Week article, Caradonio pointed out that the Abby Kelley Foster charter school has no LEP students (compared to 7 percent in Worcester public schools) and only half the proportion of free-lunch eligible students as the traditional Worcester public schools (Schnaiberg 1999).

We should note that while our focus in this study is on traditional public schools, there is some evidence that, at least in the District of Columbia, charter schools feel that they are “being punished” by the traditional district leadership. We noted earlier that the charter school leaders in D.C. have frequently charged the DCPS central administration with harassing them over the issue of buildings. And the schools chartered by the Board of Education complain of lack of administrative support from the DCPS, including on such basic items as student records (Henig et al. 1999). In fairness, Henig et al. argue that not all of the problems can be chalked up to bureaucratic hostility: “some of the tension is a predictable and perhaps unavoidable offshoot of the deliberately ambiguous nature of the charter concept. . . . From the standpoint of DCPS, there is concern that those fundamentally hostile to public bureaucracies will use charters to embarrass them and to dissipate resources . . . fears that the charter movement might precipitate the unraveling of the traditional public education system are deeply held and not without foundation.” (1999, 78).

While hostility is common, there are some examples of cooperation. The best example we found was in Springfield. In that district, which uses controlled choice to achieve desegregation that was court-mandated, the charter schools are part of the system; while charter students are selected by lottery, the charter schools agreed to be part of the controlled choice system to achieve racial balance. Again, this may not be surprising given the positive view of charters held by district leaders and the institutional links between schools in the two sectors.

Finding: Public School Officials Do Not Believe The Charter Schools Actually Provide New Models Or Programs, Limiting Their Impact

While there may be a fundamental hostility between the two sectors that leads to a lack of communication and sharing, there is also evidence that officials of traditional public schools do not believe charter schools in these cities are beacons of innovation, particularly in terms of curricula. Thus, even if lines of communication between the sectors were open, in reality public school officials may not want to listen.

Clearly, many charter school innovations are based on a return to basics (which Jersey City school officials dismiss as “rote” education) and a rejection of “progressive” education that may be more popular in some of the regular public schools. Even in Trenton, where Superintendent Lytle embraces competition from the charter schools, he argues that they are not particularly innovative, emphasizing instead basic education and parental involvement.

In Worcester, Superintendent Caradonio, teachers’ union president Brophy, and the district research director all believe that the charters in their district have not introduced any new educational ideas. According to Caradonio, while they do offer extended day programs that parents want, the rest of their package is the recycling of old, unproven ideas, and “selling snake oil.”

Even in Springfield, where Superintendent Negroni and several members of the elected school board are positive about the effects that charters are having on the system in general, none of them point out any particular innovations in the charter schools, apart from before- and after-school programs.

Districts recognize some exceptions to their beliefs. For example, in Trenton, the Granville/Edison Project charter school introduced laptops for all students and their parents—in response, the district is now examining the development of “net-schools” which would give parents laptops that could be easily connected to a mainframe controlled by the school.

But for the most part, charter schools define innovation differently than do public school officials; they want to please their “customers.” In Worcester, Steven Wilson, CEO of Advantage schools argues that his schools, like Worcester’s Abby Kelley Foster, are responsive to their “customer base.” “These parents,” he states, “crave a school setting that is orderly and safe and focused and on task. And that’s the brand we endeavor to provide them with.” (Schnaiberg 1999). Of course, the fact that these ideas are considered innovative by charter school operators and that parents enroll their children in these schools is itself a significant indicator of what the public finds lacking in their district public schools.

Finding: Charter Schools and Traditional Public Schools Differ in the Pattern of Innovations They Adopt

This difference in outlook leads to significantly different emphases on which innovations each sector is likely to adopt.

In our survey of school principals, we presented principals with a list of common innovations that schools throughout the nation are implementing and we asked which of these reforms they have introduced or expanded in the last year. We presented the charter school principals in D.C. with the same list and asked them the same question. In Figure 7, we compare the pattern of innovation across the schools in these two sectors.

As is evident in our data, we found that the regular public schools are much less likely to have implemented or expanded any of the reforms in our battery than are charter schools.14  It is interesting to note that the innovation most likely to be adopted by the traditional public schools is the creation or expansion of programs that take place before or after the traditional school day. In our fieldwork, the expanded school day was the aspect of charter school programs that was most frequently mentioned by traditional public school leaders as attractive to parents. We believe that this is probably the single clearest case of charter schools leveraging school-level change in the traditional public schools that we have uncovered in our studies.

We should note two further dimensions to this issue. First, we found that while charters often adopt many of these innovations as part of the definition of the school culture and mission, traditional public schools usually adopt reforms in a piecemeal fashion and do not use them to redefine their mission or fundamental style of management. Thus, charter schools are more likely to adopt a wider suite of reforms and try to integrate them into the culture of the schools, while traditional public schools are more likely to adopt single reforms, such as expanded day programs, without changing their standard operating procedures.

Second, at least one of the reforms we queried school leaders about may be particularly difficult for traditional public schools to implement. We found that many of the academic leaders we talked with (especially principals of traditional public schools) said they would like to involve parents more in the day-to-day operation of the schools, but that they cannot negotiate the kinds of parent involvement contracts that are common to charter schools. Since increasing parental involvement may be particularly important in creating a viable school community (see, e.g., Schneider, Teske and Marschall 2000), the limits on the ability of traditional public schools to use such parental contracts may handicap them relative to the charter schools.

Finding: Superintendents and Principals In Traditional Public Schools Perceive Limits In Their Freedom To Undertake Action In Response to Competition—But This May Be Changing

While we believe that the differences between the two sectors are often exaggerated by school officials, it is clear that the traditional public schools are more constrained by state rules and union contracts in their ability to enact reforms. The example of parental involvement contracts we just discussed is one illustration. The issue of teacher pay and teacher assignments is another critical element.

For example, Worcester Superintendent Caradonio admitted that charters offer the before and after school programs that parents want and that his principals are trying to do the same, but are hampered by the requirements of the union wage scale. Union president Mark Brophy confirmed this constraint and added that the charters bypass the problem by paying what he characterized as “less than a living wage” and by not paying overtime for extended days. The president of the Springfield union, Tim Collins, noted that the statewide Massachusetts Teachers’ Association had lobbied to create the Horace Mann category of charter schools and to get most of the union contract to apply to teachers in these schools. At the Horace Mann schools there are extended day programs, but teachers are compensated for this extra commitment—and this drives up the cost of these programs.

In New Jersey, where teachers in charter schools need to be certified, teachers’ union officials note that most of the teachers in charter schools have joined the unions. This raises a question of whether future negotiated bargaining will lead charter school teachers to question the longer hours that they now work, compared to teachers in regular public schools.

Traditionally, one limiting factor on reform in public schools has been the inability of superintendents to replace failing principals who, as front-line managers, are key to the success of schools. However, we found that in several of the districts we studied, superintendents have replaced a significant number of principals, helping to expand the number of entrepreneurial leaders in the district and helping to create a “management team” more interested in reform. In some districts, charter schools help provide the justification for such turnover—but since charter schools came on line at the same time as other changes that increased the ability of superintendents to fire principals a causal link is hard to prove.


Parents as “consumers” exercising choice are a major force for change, and as consumers parents are clearly “voting with their feet” in favor of charter schools. Given growth of charter school enrollments, it is not surprising that public school officials know it.

Peter Levanos, who advises parents in Springfield, notes that many parents know of the Sabis International School’s reputation for improved test scores. Trenton teachers’ union president Kevin O’Brien notes that both parents and school officials are well aware of the presence of competition. Officials in Jersey City believe that parents like the newer buildings and the smaller schools that characterize charter schools there.

But even parents who do not enroll their child in a charter school may gain more leverage from the presence of such schools. We argue that there is a relationship between exit and voice—as parental exit options increase, schools are forced to pay more attention to the preferences (voice) of parents. For example, in Trenton, Lytle notes that in neighborhoods most affected by the new charter schools two elementary schools chose to expand their grades from K-6 to K-8, because parents expressed great concern about the level of safety in the regular public middle schools.

This added parental voice may ultimately pressure all public schools to become more “consumer friendly.” But in the next section, we show that, at present, there is a large gap between charter schools and traditional public schools, at least in the District of Columbia.

Finding: Charter Schools Are More Consumer Friendly: Evidence from the District of Columbia

To assess the level of “consumer friendliness” of charter schools and traditional public schools, we sent a set of parents to visit a sample of Washington D.C. schools (both public charter and traditional public schools). We asked them to report on the condition of the schools they visited, how well they were treated, and how responsive the school staff was to their requests for information about programs and performance.15

Results of Parent Visits: Physical Conditions

Figure 8 shows substantial differences in the physical conditions of the DCPS schools and the public charter schools. On all four dimensions measuring the physical environment, DCPS schools are in worse condition than the charter schools. For example, our site visitors found slightly more than 10 percent of the DCPS schools had broken windows and more than 20 percent were marked by graffiti. In contrast, none of the parents visiting the D.C. charter schools found these conditions. Moreover, compared to the charter schools, almost twice the proportion of DCPS schools had bars on their windows and while almost one-quarter of the DCPS schools were located on streets with abandoned buildings, no charter schools were on streets with abandoned buildings.

There is a body of research linking these conditions to feelings of safety, actual levels of safety, and performance (Schneider et al. 1999). Our visual evidence points to the fact that the charter schools provide a better environment than the DCPS schools.

These conditions are external to the school and may be a product of a variety of factors over which the school itself has only minor control, including, for example, the actions of building inspectors or the age of the neighborhood. But we also asked parents to evaluate a set of physical conditions that are more under the control of the school. As demonstrated in Figure 9, we again find a consistent set of results indicating better performance by the charter schools—their facilities were more likely to be judged in good working order; they felt cleaner, safer and more exciting.

How Do Schools in Different Sectors Treat Parents?

As part of this research, parents entered the schools and asked staff about the schools’ programs and performance. We asked our parents to evaluate how well the staff treated them. As is evident in Figure 10, charter school staff were evaluated as being both more responsive and more courteous than the staff of the D.C. public schools visited.

In addition, we also asked parents to evaluate how responsive staff was to a series of requests for information about school programs. As shown on the following page in Figure 11, on each measure that we report, the charter schools were by far more responsive to the legitimate requests of parents for information essential to judging the appropriateness of a school for a prospective student.

We recognize that our findings are based on small numbers (although our parents did visit almost half the charter schools that were open at the time of this research and 15 percent of the traditional public schools). We also recognize that because the charter schools are relatively new they may be more likely to be located in newer facilities located in better locations. And it may be that their relatively smaller size enables their staff to be more parent-friendly. Nonetheless, the patterns here are strikingly consistent. At this time, both the “hardware” of the charter schools (the buildings and their facilities) and the “software” (the staff) are consistently more conducive to an environment supportive of a good educational experience and more appealing and open to parents.


There are two other issues that we feel are central to studying the relationship between charter schools and the traditional public schools. The first one focuses on the academic performance of charter schools and traditional public schools, and it has been discussed at some length in national debates. The second issue we believe has been under-studied but may emerge as a central issue as charter schools continue to grow—the movement of children from private schools to charter schools. While we do not have extensive empirical evidence on the extent of this movement, we do find anecdotal evidence in most of the districts that this flow is becoming increasingly common.

Finding: There is Little Evidence At Present Relating Test Scores in Charters andRegular Public Schools

Due to lack of systematic data over time, there is as yet no firm evidence of higher performance by students of charter schools.

Nevertheless, in Springfield there has been considerable “buzz” among officials and parents that the Sabis school has improved test scores. There is evidence supporting this argument: The Center for School Change at the University of Minnesota reports that when the school was converted from a regular public school, in 1995, only 38 percent of students performed at or above grade level, making it one of the lowest-performing schools in Springfield (Cheung et al. 1998). At the end of the second year, with essentially the same student body, about 62 percent of Sabis students performed at or above grade level. Superintendent Negroni used this evidence of higher performance in his own campaign to improve the rest of the Springfield schools.

The success of Sabis forced the traditional public schools to put much more emphasis on testing their students more often, according to Peter Levanos, head of Springfield’s interdistrict choice program. Several schools also started up Saturday programs, to match the Saturday programs at Sabis. Levanos noted that this competition has led to ten of Springfield’s twenty elementary schools now outscoring Sabis, five of which were previously scoring at those levels, with the other five stepping up to “beat Sabis.”

In Worcester, Research Director Patty Mostue stated that out of 34 elementary schools in the city, the two charter schools are scoring in the bottom ten. The district did not, however, report whether scores at the two charter schools were better or worse than at traditional public schools serving children from similar demographic backgrounds.

In New Jersey, there is not yet any good information on charter school test scores and whether or not they are influencing the regular public schools. Test scores in Jersey City have improved over the period of the state take-over and they are higher now than in Trenton. This provides fuel for the Jersey City approach of proceeding with their own reforms, rather than actively competing with the charter schools.

Finding: Charter Schools May Be Evolving as Substitutes For Private Schools

We have identified an unanticipated effect of charter schools on private school enrollments. Some parents seem to perceive charter schools as “free” private schools, in part because, like many private schools, they are small, safe, and often have thematic emphases. Many charter school pupils are coming from parochial schools, particularly in Trenton, where the local Catholic diocese is closing several schools, some of the physical facilities of which are being taken over by the regular public schools and by the charter schools.

While there are little firm data, Trenton’s Lytle estimates that as many as 30 percent of the charter school students are former private school students. According to the Trenton Times, three Catholic schools in Trenton have been closed since 1993 and the diocese claims to have lost 450 students, representing 15 percent of its total enrollment, since 1996, when charter schools were approved. Diocesan spokesperson Eileen Marx notes about the role of charter competition in these losses: “It is certainly a factor. It’s a big piece.” (Southwick 1999).

Similarly, in Jersey City, as many as twenty percent of charter school students may be coming from private schools (of course it is impossible to determine at the Kindergarten level where parents would have sent their child absent the option of charter schools). While this is part of a long-term decline, Jersey City Catholic school enrollment fell 2 percent from 1995 to 1997, but fell 5 percent in the next two years, after charter schools entered the environment (Parello 1999). Overall, in 1998-9, N.J. state statistics showed that 18 percent of charter school students across the state had previously been in private schools. Not surprisingly, the number is higher in urban areas.

This student flow in New Jersey suggests that charter schools are creating a larger market for schools, expanding the choice set of parents who might otherwise have left the public school sector entirely for sectarian schools. For some observers, this is a positive development, as the retention of students (and their parents, who are clearly motivated to make active choices) in the public school system has many positive benefits. For others, such movement suggests that charter schools are “subsidized” private schools for parents who would otherwise not cause any financial expense to the public system.

This student flow seems to be less of an issue in Massachusetts, where a state Department of Education report claims that 96 percent of charter school students came from public schools, with only four percent coming from private schools (Massachusetts Department of Education 1998). But even in Massachusetts, there is some evidence that parochial schools are responding to this new form of competition. In March 1999, the Boston Archdiocese launched its first-ever television advertising campaign in support of Catholic education. Church officials claim, however, that this campaign was not a response to the threat of the new charter schools; rather, the ads simply represented the next stage in a long-term strategy to reverse “decades of declining enrollment.” (Lazar 1999).

Finally, although no official documentation exists, movement from private schools into the public charter system in the D.C. appear to be closer to the experience in Massachusetts. In our telephone survey, 18 D.C. charter school principals report that at least 90 percent of current charter students were believed to have previously attended traditional public schools, but they also estimated that between 0 and 10 percent of their students previously attended private schools. Although the effects of charter substitution for private schools are relatively small at this point, the availability of public alternatives to private education via new charter schools each year will continue to increase. Undoubtedly, many parents who exited the traditional system in search of private sector alternatives will find it increasingly in their interest to explore the potential of the publicly funded charter system.


While competitive effects from charter schools are not as extensive as many advocates had hoped for, our research has uncovered some reasons why this may be so: resistance from officials who don’t like charter schools; policies that shield districts and schools from feeling the financial consequences of failing to attract students; increasing enrollments because of demographic changes; a belief among traditional public school leaders that charters are not doing innovative things that are worthy of imitation. Our research also uncovered evidence that traditional public schools are reacting to charter school competition when they have reason to feel threatened. Principals report they change educational and administrative procedures when they feel competitive pressure, and districts adopt new programs when they can clearly see that parents want those programs. Those who believe traditional public schools can be driven to improve efficiency and achievement from competition can take heart from these findings.

States wishing to see more competitive effects ought to consider changing many of the policies that currently shield public schools from feeling the effects of competition and hinder their ability to respond when they do feel pressured. These changes would include:

  • Enacting and implementing statewide school finance policies that subtract money from school districts on a per-pupil basis when students enroll in charter or other non-district schools;
  • Resisting the temptation to replace any financial losses caused by these policies through increases in other funding sources for affected districts;
  • Encouraging districts to enact and implement similar per-pupil financing policies for individual schools;
  • Providing principals with more autonomy to effect administrative and educational changes at the school level; and
  • Encouraging districts to work with teachers’ and other unions to foster experimentation and flexibility in educational and administrative practices.

States that embark on such reforms will provide the conditions for a purer test of the effects of market-style competition on schools and school systems, enabling parents and policy makers to determine whether such competition really does improve student achievement or produce other measurable improvements in the quality and efficiency of the educational experience.

Our research also points to lessons for those leading traditional public and charter schools. Traditional public school leaders need to start seeing school policy decisions from the viewpoint of parents. Clearly, many parents like what charter schools are offering, and, based on our data from D.C., it is easy to see why. Maybe what charter schools are offering does not appear innovative to many public school administrators, but to parents and charter school operators it can be a novel and well-delivered blend of basic elements now lacking in the public schools.

Charter school operators can also learn to be more forthcoming and communicative about what they are doing. Competition can lead to good outcomes, but everyone recognizes that education is different from other consumer goods in at least one respect: children get only one chance to become educated. If traditional public school leaders and charter school operators can learn to communicate better, even as they remain competitive, the worthwhile lessons from the charter school experiment will spread much more quickly. That can only benefit our children.



We contracted D.C. Parents for School Choice, a Washington-based non-profit organization, to train parents to visit both public charter and D.C. public schools. These parents were trained to pose as parents thinking about enrolling their child in the particular school and were trained to ask for information pertinent to that enrollment decision. Since all of the parents selected had recently enrolled their own children in new schools, they were already familiar with enrollment procedures and what to look for in schools.

We prepared a checklist for each parent to fill out reporting on the observed conditions in the schools they visited and reporting on what they experienced during their visit. Parents were asked to report on two different aspects of their visit. First, they were asked to evaluate the physical condition of the school and its neighborhood, and, more subjectively, to evaluate the “feel” of the school. Second, parents were instructed to ask school personnel for information about the programs that the schools offer and to ask for materials about programs and performance. These second types of data were gathered to assess differences between schools in the degree to which school staff was responsive to requests for information that is important in allowing parents to evaluate the appropriateness of the schools for their children.

Traditional D.C. public schools visited:
Amidon ES
Anne Beers ES
Brightwood ES
Brookland ES
Cook ES
Davis ES
Lafayette ES
Langdon ES
Mildred Green ES
Miner ES
Randle Highlands ES
Backus MS
Brown MS
Browne JHS
Jefferson JHS
Lincoln MS
Ballou SHS
Coolidge SHS
Eastern SHS
School Without Walls SHS

D.C. public charter schools visited:
Chamberlain (K-5)
Children’s Studio (ungraded, ages 3-8)
Community Academy (K-3)
Edison-Friendship (K-5)
Options (grades 5-8)
The SEED school (grades 7-8)
Village Learning Center (K-8)
Cesar Chavez PCHS (grades 9-10)
IDEA academy (grades 9-12)
Richard Milburn (grades 9-12)
The Washington MST (grades 9-12)


In this paper, we built on existing work to study how charter schools influence the traditional public school system. Given the early state of research into this question, as well as time and budgetary constraints, we chose to undertake relatively detailed case studies in a small number of school districts rather than to try to construct a large national database. Our goal was to increase our understanding of the extent to which both individual schools and urban districts as a whole respond to charter competition and to identify the mechanisms that either impede or facilitate that response. We chose to study five urban districts facing growing competition from charter schools: Springfield, Massachusetts; Worcester, Massachusetts; Jersey City, New Jersey; Trenton, New Jersey; and the District of Columbia (D.C.).

We had several criteria in mind in choosing these cities. First, we chose to focus on urban districts because it is in those districts that so many of the problems common to the public education system are manifest. The need to address these problems has been increasingly addressed by the introduction of competition to the traditional school system via charter schools.16  Second, we chose to focus on mid-size urban districts because they were more accessible than the largest cities for the type of exploratory study we conducted. Third, we chose not to study cities in the states in which the spread of charters is most advanced (e.g., Arizona, Michigan), since these locations are already receiving scholarly attention. Conversely, we avoided states that only recently passed charter legislation (e.g., New York), as the impact of charters on the traditional schools is likely to be small and too early to detect. However, we wanted to insure that the cities selected for study would be representative of the widely varied strengths of charter legislation throughout the country. The laws in Massachusetts and D.C. have been ranked among the most favorable to charters, while New Jersey’s law is ranked in the middle of the 37 states studied by the Pacific Research Institute (Billingsley and Riley 1999). Finally, in order to “hold constant” some of the situational context in which cities are attempting to meet the challenges of education reform, we chose two cities in each of two states.

While we acknowledge that our cases do not represent a random sample of urban school districts, the results we report are congruent with theories predicting the manner in which competition can affect traditional public schools (see, e.g., Schneider, Teske and Marschall 2000; Chubb and Moe 1990). Our study also indicates several new patterns in the relationship between charter schools and traditional public schools beyond those documented in existing studies.

A Brief Introduction to the Districts

As noted above, we studied the effects of charter competition in five urban districts. The two Massachusetts districts were Springfield and Worcester, both with approximately 26,000 students. In New Jersey, we studied Trenton (13,500 students) and Jersey City (33,000 students). The fifth city, Washington, D.C. (77,500 students), is a larger district that has by far the greatest number of charter schools of districts in our analysis. In the following section, we briefly outline the conditions found in these districts.17 

The Massachusetts Districts

Similar to schools in many urban areas, the public schools in Springfield have experienced many problems over the past few decades. About 10 years ago, a reform coalition of Springfield business leaders and politicians decided it was essential to improve the schools, and thereby help the future development of the city. The school board, which is an elected body headed by Springfield’s mayor, hired a reform-oriented superintendent, Peter Negroni, who had been a superintendent of a New York City community school district. Negroni was clearly expected to shake up the Springfield school system and he has done so, replacing a number of principals who were not performing well and instituting an environment that emphasizes reform. When he arrived, the district was under a court order to desegregate; with Negroni’s support, Springfield has used controlled interdistrict choice as a means of voluntary desegregation.

Springfield faces competition from three charter schools in or near the district. The Sabis International School is run by a multinational corporation and is the largest of the three, enrolling over 800 students in grades K-9. North Star Academy is a storefront high school with about 140 students run by Learning Tree, an education advocacy group. The New Leadership School is a “Horace Mann” charter school resulting from a partnership between the Urban League of Springfield, the Massachusetts National Guard and the Springfield School Committee. The school opened in the summer of 1998 with grade 7 and is expanding to grade 12. It currently enrolls 75 students and plans to have around this number of students per grade.

In Worcester, the public schools have generally performed better than those in the other districts we are studying, at least partly because the student body is not as low-income as in the other cities (see Appendix 1). Thus the Worcester schools have not experienced the sense of “crisis” that has been a hallmark of the school systems in all of the other cities we studied. In fact, in recent years, as the high-technology development around the Boston area has pushed out even further toward Worcester, Worcester’s public schools have received good press and have been cited as an attraction for families considering living in the city. As a consequence of this sense of success, leadership has been fairly stable in Worcester. Three years ago the elected school board appointed James Caradonio as superintendent, who had formerly been the Deputy Superintendent of the district for several years prior to his appointment.

There are two charter schools in the Worcester area, Seven Hills, which is associated with the Edison Project, and the Abby Kelley Foster school, which is run by Advantage Schools. Abby Kelley Foster was featured in a December 1999 front page article in Education Week (Schnaiberg 1999), and one theme of the article was the fact that district officials are not supportive of the charter schools—a point we discuss in this report.

The New Jersey Districts

In Trenton, for the past decade, the schools have performed at the very lowest level in the state, even when compared to other low-income urban areas. The district barely escaped being taken-over by the state 10 years ago when several other urban systems became state-run operations. Under the previous chief schools administrator, the Trenton schools were clearly failing and the district’s efforts at reforming operations were lagging. As a consequence, in 1998 the Mayor and the appointed school board hired James (“Torch”) Lytle, who was known as a reformer from his experience in Philadelphia—and was also known as a supporter of charter schools. Indeed, while in Philadelphia, Lytle had even written a charter school proposal. His approach, like that of Negroni in Springfield, has been to shake things up quickly. He noted: “Our kids have been failing state tests and dropping out for too long. We can’t pretend that tinkering is going to make things better. . . . We need to transform schooling if we are to get where we need to be.” As an example of his efforts, and despite some controversy, he is breaking down the large and traditional Trenton High School into several smaller, thematic “learning communities.” (Coryell 1999).

There are seven charter schools in the Trenton area, of which the Granville School, which is a partner with the Edison Project, is by far the largest. The Village Charter School, with a state-of-the-art new building, is the second largest. Others include the International Charter School, the Greater Trenton Academy and Technology School, the Emily Fisher School, the Trenton Community Charter, and the boarding school Proctor Academy.

Jersey City schools also performed poorly but, in contrast to the Trenton schools, were actually taken over by the state 10 years ago, partly because there was also corruption related to the elected school board, which was using the schools to provide patronage employment. The current superintendent, Richard DiPatri, was a state Education official appointed three years ago to head the Jersey City schools. DiPatri has undertaken several reforms, including appointing a new supervisor of principals, to try to make principals more accountable to the system. In the last few years, student test scores in Jersey City have improved to nearly the best among the 12 most low-income urban areas in the state and the district is now meeting most of the state monitoring indicators (it was meeting only 37 percent of them 10 years ago). As a result of this improvement, the state plans to cede authority for Jersey City’s schools back to the elected Board in two years.

Mayor Bret Schundler, an emerging Republican political star, strongly supports vouchers and school choice, and city school board elections largely involve the Mayor’s candidates squaring off against those of the teachers’ union. While Schundler believes the state improved the schools, in part because they increased the district’s annual budget from $180 million to $380 million, he notes that, “…less than half our kids are graduating from high school and there has been only a small increase in test scores.” (Newman 1999).

There are nine charter schools in the Jersey City area (seven in Jersey City and two in nearby Hoboken), of which the Golden Door School is the most prominent. Others include the Gateway School, the Second Chance High School, Jersey City Community Charter, Learning Community Charter, Soaring Heights Charter, and the Elysian Charter.

The Washington D.C. Schools

The public school system in the District of Columbia has long been regarded as one of the worst in the country, despite funding that is 26 percent higher than the national average. A 1996 report (Board 1996) documented system-wide deficiencies in areas ranging from basic operation—for a variety of reasons, many tracing back to mismanagement, D.C. schools had not opened on time in three years—to extremely poor test scores—70-90 percent of tenth graders were consistently scoring below national averages in reading and math.18  As a result, the D.C. Financial Control Board declared a state of emergency in the schools and seized control. Using a military crisis approach, the Board hired three-star Army General Julius W. Becton Jr. as superintendent, who was replaced in 1998 by the current superintendent Arlene Ackerman.

The 1996 federal law allowing charter schools in D.C. vested chartering authority in the elected D.C. Board of Education and in a newly created (and appointed) Public Charter School Board. There are now 29 charter schools in D.C., and they serve one out of every eleven public school students, with an enrollment of approximately 6,700, compared to the almost 71,000 students enrolled in the traditional public schools. D.C. may represent the fastest growing charter system in the country, and there should be enough capacity for one-quarter of D.C. students to be enrolled in charter schools by 2004.

In Appendix 1, we provided a table with summaries comparing several indicators of conditions found in the five districts.


  • Barzelay, Michael and Babak J. Armajani. 1992. Breaking Through Bureaucracy: A New Vision for  Managing in Government. Berkeley: University of California Press.
  • Billingsley, K. Lloyd and Pamela Riley. 1999. Expanding the Charter Idea: A Template for Legislative and Policy Reform. San Francisco: The Pacific Research Institute for Public Policy.
  • Board, District of Columbia Financial Control. 1996. “Children in Crisis: A Report on the Failure  of D.C. Public Schools.” The District of Columbia Financial Responsibility and Management Assistance Authority.
  • Byrk, Anthony S., Valerie E. Lee, and Peter B. Holland. 1993. Catholic Schools and the Common Good. Cambridge: Harvard University Press.
  • Cheung, Stella, Mary Ellen Murphy, and Joe Nathan. 1998. “Making a Difference? Charter Schools, Evaluation and Student Performance.” Center for School Change, Minneapolis, MN.
  • Coryell, Lisa. 1999, December 9. “Superintendent Talks About Future of Trenton’s Schools.” Times  of Trenton, Trenton, NJ. p. A3.
  • Hale, Sandra. 1996. “Achieving High Performance in Public Organizations.” in Handbook of Public Administration, 2nd ed., edited by J. Perry. San Francisco, CA: Jossey-Bass.
  • Henig, Jeffrey R., Michele Moser, Thomas T. Holyoke, and Natalie Lacireno-Paquet. 1999.
     Making a Choice, Making a Difference? An Evaluation of Charter Schools in the District of Columbia. The Center for Washington Area Studies, George Washington University, Washington, D.C.
  • Hess, Frederick M. 1999. “Hints of the Pick-Axe: the Impact of Competition on Public Schooling  in Milwaukee.” Association for Public Policy Analysis and Management, Washington,  D.C.
  • Kolderie, Ted. 1990. Beyond Choice to New Public Schools: Withdrawing the Exclusive Franchise  in Public Education. Washington, D.C.: Progressive Policy Institute.
  • Lazar, Kay. 1999, March 7. “Parents Forsake Public for Private, Parochial Schools.” Boston Herald, Boston, Mass. p. 1.
  • Massachusetts Department of Education. 1998. Massachusetts Charter Schools: Expanding the Possibilities of Public Education. Boston, Mass.: Massachusetts Department of Educa- tion.
  • Nathan, Joe. 1996. Charter Schools: Creating Hope and Opportunity for American Education. San  Francisco, CA: Jossey-Bass.
  • Newman, Maria. 1999, May 27. “State Plans Return of Jersey City Schools to Local Control.” New  York Times, New York, NY. p. B6.
  • Osborne, David and Ted Gaebler. 1992. Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector. Reading, Mass.: Addison-Wesley Pub. Co.
  • Parello, Nancy. 1999, March 20. “Charter Schools Stealing Catholic School Students.” Times of Trenton, Trenton, NJ. p. B5.
  • Purkey, S.C. and M.S. Smith. 1983. “Effective Schools: A Review.” The Elementary School Journal 83:427-54.
  • Rosenblum Brigham Associates. 1998. Innovation and Massachusetts Charter Schools. Boston,  Mass.: Massachusetts Department of Education.
  • Rothstein, Richard. 1998. “Charter Conundrum.” The American Prospect, pp. 46-60.
  • Schnaiberg, Lynn. 1999. “Seeking a Competitive Advantage.” Education Week, December 8, p. 1.
  • Schneider, Mark, Melissa Marschall, Christine Roch, and Paul Teske. 1999. “Heuristics, Low Information Rationality, and Choosing Public Goods: Broken Windows as Shortcuts to In- formation about School Performance.” Urban Affairs Review 34:729-741.
  • Schneider, Mark, Paul Teske, and Melissa Marschall. 2000. Choosing Schools: Consumer Choice and the Quality of American Schools. Princeton, NJ: Princeton University Press.
  • Shoemaker, Crissa. 1998, November 3. “Expectations Are Great for Charter School.” Times of Trenton, Trenton, NJ. p. A2.
  • Southwick, Ron. 1999, March 3. “Blessed Sacrament School to Close in Trenton.” Times of Trenton, Trenton, NJ. p. A3.
  • Strauss, Valerie. 1999, September 30. “Parents Seek to Save School; D.C. Officials Order Chartless  Kwame Nkrumah to Close Today.” Washington Post, Washington, D.C. p. B 02.
  • Strauss, Valerie. 2000, January 10, 2000. “A Leader’s Pass-Fail Test.” Washington Post, Washington, D.C. p. B1.
  • Wilgoren, Debbi. 1998, December 18. “Chavous Proposes New Office to Supervise Charter  Schools.” Washington Post, Washington, D.C. p. C 04.


  1. We call these “regular” or “traditional” public schools, to distinguish them from charter schools, which are also publicly funded.
  2. We present a fuller description of our methodology and a description of each district’s background in Appendix 3.
  3. The classic discussion of adaptive public sector organizations with a customer orientation is Osborne and Gaebler’s Reinventing Government (Osborne and Gaebler 1992). Other relevant discussions include Hale (1996) and Barzelay and Armajani (1992).
  4. Note that even in districts with the strongest support from the superintendent response rates, while high, were not 100 percent. We believe there is a simple lesson here: while the support of the superintendent will help generate principal participation, it will not guarantee full participation. On the other hand, opposition from a superintendent can make access to principals virtually impossible.
  5. We recognize that there may be a selection bias affecting the pattern of principal responses. While there is no relationship between the level of support given to charters and the level of effort the superintendent put into getting principals to respond, the principals who did respond could represent a non-random set of the total population of principals.
  6. Unless otherwise cited, all quotes are taken from our interviews with the person quoted.
  7. Massachusetts General Law Chapter 71 § 89.
  8. It is likely that this will change as the scheduled reductions take place. Indeed, according to Worcester Superintendent Caradonio, his district, presently with 1,100 charter pupils, could lose up to $5 million in three years when the reductions begin to more fully reflect the recent growth in charter school enrollment.
  9. 153 N.J. 480.
  10. Our finding of blunted financial losses is not limited to these case districts. Hess (1999a) found that the Milwaukee public schools received increased funding to make up for the financial losses from students choosing private schools with vouchers.
  11. Of course, district superintendents do not hire themselves. While we emphasize the role of superintendents, we recognize that their appointments and some of their success is related to their relationships with the school board and local politicians. Nevertheless, superintendents have a great deal of freedom in the operation of their school systems and how they respond to charter school competition is not necessarily dictated by local officials (see, e.g. Hess 1999b). In D.C., the relationship of the superintendent to other political officials is even more complex, as Ackerman must deal with several elected and appointed officials in D.C. and in Congress.
  12. For a detailed recent profile of Ackerman’s leadership style, see Strauss (2000).
  13. The “pressure” variable is a count of the number of competitive sources an individual principal expects to lose students to in the near future, including charter schools, private schools, and intra-district movement of students to other public schools.
  14. There are two limitations on these data that must be noted. First, there is the obvious problem that we are comparing traditional public schools in districts other than in D.C. to charter schools in the District. Second, there is the problem of the “baseline”—that is, some innovations may not be adopted or expanded in some schools because they are already in place for some time. In our analysis, we tried to minimize this problem by dropping from consideration reforms that were already in place in most traditional public schools and which could not be expanded. For example, we dropped all day kindergarten from this chart because most public schools in our survey already had them and it is a “binary” condition—either you have it or you don’t have it—and, thus, asking principals if they “expanded” the program makes no sense. In contrast, we believe that the innovations that do appear in Figure 7 can be expanded or else were in place in relatively small numbers of traditional public schools.
  15. In Appendix 3, we describe our procedures in more detail.
  16. While schools in rural areas often have severe problems, their relatively sparse populations often limit the number of competitive schools that can be supported. And while there are many failing suburban schools, most seem to be producing relatively a satisfactory education for their students.
  17. Our goal in this report is not to provide a comprehensive history of the development of charter schools in these cities. Rather, we are outlining the basic institutional features in these cities and the manner in which charter schools affect the behavior of the three sets of stakeholders—superintendents, principals and parents—with which we are most concerned in this study. For fuller explorations of the development of charters in these states: see, for example, Henig et al (1999) on D.C., Kane (1998) on New Jersey, and Massachusetts Department of Education (1998) on Massachusetts.
  18. One other basic task on which the DCPS has failed (and which affects some of the numbers we report in Appendix 1) is accurately counting their students. The 1996 report on public education in the District of Columbia, “Children in Crisis” reports that “the DCPS does not have an accurate count of its students. Estimates vary between 75,000 and 81,000 students. The National Center for Education Statistics (NCES) found a discrepancy of 20.6% between the 1990 census and the number of students reported by DCPS. The General Accounting Office (GAO) also has questioned DCPS’ record keeping. After the 1995 sample student enrollment count authorized by the Superintendent identified 80,450 students, the GAO stated that its usefulness for validating enrollment was ‘limited because of mistakes made in selecting the sample.’ GAO estimated that DCPS’ Student Information Management System may contain approximately 5,000 obsolete or duplicative student records.” (Board 1996)


Center for Civic Innovation.



Many education reformers argue that competition for students from charter schools can force traditional public schools to improve. This pathbreaking study of five urban school districts in the Northeast, including Washington, D.C., finds only limited proof of much improvement to date.


Summary of Findings

Why Should School Districts Respond?

How New Ideas Can Leverage Change

How Financial Pressure Can Leverage Change

Overview of Our Methodology

Table 1: Principal Response Rates Vary Widely Across Districts

Several Factors Have Limited the Effects of Charter Schools on Traditional Public Schools

Finding: District-Level Attitudes Toward Charter Schools Vary Widely

Figure 1: Levels of Support Vary Across Districts

Finding: District-Level Support is Not Related to Market Share Held by Charter Schools

Figure 2: There is No Evident Relationship Between Level of Support and Percent of Students Enrolled in Charter Schools

Finding: School Districts Have Been Shielded from the Fiscal Implications of Losing Students

Finding: To Date Individual Schools Have Been Held Harmless

Finding: Population Trends May Also Be Blunting the Impact of Charter Schools

Figure 3: Charter School Students Represent a Growing Share of D.C. Enrollments

What Factors Affect How Traditional Public Schools Respond?

Finding: The Role of School Leadership is Critical in Reform

Finding: Even if Districts Remain Insulated, Principals Respond to Pressure

Figure 4: As Competitive Pressure Increases, Principals Innovate More

Figure 5: As Competitive Pressure Increases, Principals Spend More Time on Increasing Efficiency

Figure 6: As Competitive Pressure Increases, Principals Feel They Don’t Have Enough Autonomy

Perceived and Actual Differences Between Schools in the Two Sectors Limit the Leverage That Charters Are Exerting

Finding: There is Hostility Between the Sectors that Limits the Spillover from Charter Schools to Traditional Public Schools

Finding: Public School Officials Do Not Believe the Charter Schools Actually Provide New Models or Programs, Limiting Their Impact

Finding: Charter Schools and Traditional Public Schools Differ in the Pattern of Innovations They Adopt

Figure 7: Traditional Public Schools and Charter Schools Differ in the Range of Innovations They Adopt

Finding: Superintendents and Principals in Traditional Public Schools Perceive Limits in Their Freedom to Undertake Action in Response to Competition—But This May be Changing

Charter Schools Treat Parents Better than Traditional Public Schools

Finding: Charter Schools are More Consumer Friendly: Evidence from the District of Columbia

Figure 8: Differences in the Physical Environment of D.C. Public Schools and D.C. Charter Schools

Results of Parent Visits: Physical Conditions

How Do Schools in Different Sectors Treat Parents?

Figure 9: Other Conditions in D.C. Public Schools and D.C. Charter Schools

Figure 10: How Well Did Staff Treat Parents?

Figure 11: Additional Indicators of How Parents Were Treated

Other Issues

Finding: There is Little Evidence at Present Relating Test Scores in Charter and Regular Public Schools

Finding: Charter Schools May Be Evolving as Substitutes for Private Schools


Appendix 1: Summary of Conditions in the School Districts Studied

Appendix 2: D.C. School Visits Procedures

Appendix 3: Approach and Methodology

A Brief Introduction to the Districts

The Massachusetts Districts

The New Jersey Districts

The Washington D.C. Schools




Home | About MI | Scholars | Publications | Books | Links | Contact MI
City Journal | CAU | CCI | CEPE | CLP | CMP | CRD | ECNY
Thank you for visiting us.
To receive a General Information Packet, please email
and include your name and address in your e-mail message.
Copyright © 2015 Manhattan Institute for Policy Research, Inc. All rights reserved.
52 Vanderbilt Avenue, New York, N.Y. 10017
phone (212) 599-7000 / fax (212) 599-3494