Nationally, charter school students are more likely to be from families living in poverty, as measured by students’ eligibility for free or reduced-price lunch. A 2013 report from the Center for Research on Education Outcomes (CREDO) at Stanford University found that in the 2010–11 school year, 48 percent of all public school students—3.7 percent of whom were charter students in that year1 — were eligible for free or reduced-price lunch, whereas 53 percent of all charter school students were eligible for free or reduced-price lunch.2
According to the U.S. Department of Education, in 2011–12, charter school students nationwide were more likely to be African-American or Hispanic than traditional public school students. And charter school students were less likely to be white and less likely to be Asian than traditional public school students.3
This national pattern is partially explained by the fact that charter schools tend to be concentrated in large urban centers where the student population is majority African-American and Hispanic and is more likely to be living in poverty compared with students in nonurban areas. According to the National Center for Education Statistics, in the 2011–12 school year, over half of charter schools (55.4 percent) were located in cities. Just 21.2 percent were in suburban locales, 7.4 percent were in towns, and 16 percent were located in rural areas.4
When researchers compare charter school students’ demographics with those of their peers in nearby traditional public schools, a more nuanced picture emerges: The 2013 CREDO study compared charter school students with students in local “feeder” schools—that is, schools from which charter school students are known to transfer if they do not start off at a charter school—in 27 states and found that at both charters and their traditional public feeder schools, 54 percent of students were eligible for free or reduced-price lunch in 2010–11.5
The extent to which charters educate more or less disadvantaged students than their nearby traditional public schools also varies across geographic locations. For example, Caroline Hoxby and her colleagues examined New York City students and found that charter school students in 2005–06 were more likely to live in poverty than students in New York City’s traditional public schools, as measured by their eligibility for free or reduced-price lunches. But Atila Abdulkadiroglu and colleagues, using data from 2006–07, found the opposite pattern in Boston—traditional public school students were more likely than charter school students to be living in poverty.6
The public policy research institute Mathematica, in a study of nonprofit charter management organizations (CMOs) — nonprofit organizations that operate multiple charter schools—used data through 2010-11 and reported that of 11 CMOs for which they had sufficient information on their and their districts’ students’ eligibility for free and reduced-price lunch, eight served more low-income students than their host districts, and two served fewer.7 These three studies, however, did not compare charter students only with feeder school students, which some observers argue are the more relevant comparison group. For more information about charter management organizations, see the Charter School Operators section.
Local comparisons of students’ race/ethnicity typically have confirmed one national-level trend—namely, that charter schools have disproportionally attracted African-American students. CREDO reported that in 2010–11, charter school students in the 27 states they studied were notably more likely to be African-American than students in traditional public feeder schools and hence less likely to be Hispanic or white compared with students in their feeder schools.8
Similarly, Ron Zimmer, Brian Gill and colleagues—in a study published by the nonprofit, nonpartisan public policy research institute the RAND Corporation examining data from 2006–07—found that in six out of seven urban areas and states they studied, African-American students were overrepresented in charter schools compared with feeder public schools.9 New York City and Boston school studies have also reported that charter schools in these urban centers enroll a larger number of African-American students compared with traditional public schools in these areas.10
The overrepresentation of African-American students in charter schools has led observers to ask whether charter schools contribute to racial/ethnic segregation. Indeed, a number of studies have provided evidence that charter schools tend to be less racially/ethnically diverse than both the traditional public schools in their districts and the more limited group of feeder schools from which students transfer into charters.
For example, Caroline Hoxby concluded that New York City traditional public schools were more diverse than the city’s charter schools, which were educating largely African-American students in 2005–06.11 Kevin Booker and colleagues—in a study published by the RAND Corporation examining data from 2003–04—found that in Texas and California, African-American students who transferred into charters were typically leaving more diverse schools and entering less diverse charters—that is, charters that educated mostly African-American students.12 Similarly, Ron Zimmer, Brian Gill and colleagues detected a trend for African-American students to transfer to charters with more African-American students, and hence enter less diverse schools than the ones they left, in five of seven locations they studied. Transfer students from other racial/ethnic groups typically entered charters with racial/ethnic compositions similar to those in the traditional public schools they left.13 For more information about the race and ethnicity of charter school families and how they choose, see the Families section.
Charter school studies have consistently reported that students with special education needs have tended to be underrepresented in charter schools. For example, CREDO 2013 reported that in the 2010–11 school year, 8 percent of charter school students in their 27-state study were considered special education students, while 11 percent of students in these charters’ feeder schools and 12 percent of students in the traditional public schools in these states were considered special education students.14
Location-specific studies, including those focusing on New York City, Boston and Los Angeles, have also reported that charter schools enrolled fewer special education students than did traditional public schools in their districts, including those charters’ feeder schools.15 Mathematica’s research on nonprofit charter school management organizations (CMOs), using data from 2010–11, found that 18 out of 23 CMOs for which they had relevant data served fewer students who had received a special education plan prior to entering the middle school, compared with traditional public schools in host districts.16 And Mathematica’s research on charter middle schools in one CMO, the Knowledge Is Power Program (KIPP), found that students in 2010–11 who transferred into these charter schools were less likely to have received special education than their traditional elementary school peers.17
Critics have argued—and cited anecdotal evidence—that charter schools purposefully discourage families with special needs kids from applying and/or steer them away to other schools. Several other reasons have also been suggested by researchers and other observers for why charters serve a lower proportion of special education students. It is possible that parents of children with special needs believe that traditional public schools are better equipped to educate their children. They may therefore be less likely to consider a charter school. It is also possible that charter schools use different criteria for classifying students as special needs, thereby undercounting those students. However, the latter argument cannot explain why some studies, such as Mathematica’s KIPP evaluation, found that children who receive special education status in a traditional public school are less likely to transfer into a charter school.
Charter school studies have consistently reported that students with limited English proficiency are typically underrepresented in charter schools. For example, CREDO 2013 reported that in the 2010–11 school year, 9 percent of charter school students in their 27-state study were considered English-language learners, while 13 percent of students in these charters’ feeder schools and 10 percent of students in the traditional public schools in these states were considered English-language learners.18
Location-specific studies, including those focusing on New York City, Boston and Los Angeles, also reported that charter schools enrolled fewer students with limited English proficiency than did traditional public schools in their districts, including these charters’ feeder schools.19 Mathematica’s research on middle schools in charter management organizations, using data through 2010–11, found that 13 out of 16 organizations for which they had data served fewer students considered to have limited English proficiency prior to entering the middle school, compared with traditional public schools in host districts.20 And Mathematica’s research on charter middle schools in one charter management organization, the Knowledge Is Power Program (KIPP), found that students in 2010–11 who transferred into these charter schools were less likely to have had limited English proficiency than their traditional elementary school peers.21
Again, observers have suggested several reasons for why charters serve a lower proportion of students with limited English-language proficiency. Critics argue that charters purposefully discourage English-language learners from applying to their school and/or steer them to other schools. It has also been suggested that these students’ parents may have less access to information about their school options and thus may be less likely to consider charter schools. It could also be that charter schools use different criteria for assessing students’ language difficulties, or they may use different methods to help students reach English proficiency faster. However, the latter is not an explanation for studies such as Mathematica’s KIPP evaluation that find that children who were classified as having limited English proficiency in a traditional public school are less likely to transfer into a charter school. For more information about how parents choose charter schools, see the Families section.
Questions about whether charters pull more motivated and better-prepared students (and parents) away from traditional public schools have been of great concern to many observers.
To address this question, researchers have compared the prior test scores of students who transferred from a traditional public school into a charter school—or who applied to a charter school lottery—with those of their same-school peers who did not enter or apply to a charter school. Given that charter schools are schools of choice, this methodology is the most rigorous for estimating the extent to which students who choose a charter school may be better prepared to succeed academically than the students who stay behind. However, this methodology is limited in two important ways:
Moreover, there are very few studies that have enough data to compare charter transfer students with their peers in the schools they left.
With that said, the available, albeit limited, evidence does not support the notion that charter school transfer students are consistently better prepared than their traditional public school peers.
Ron Zimmer, Brian Gill and colleagues conducted a study published by the RAND Corporation using student data from 2007–08 in Chicago, San Diego, Philadelphia, Denver, Milwaukee and the states of Ohio and Texas. They found that in those metropolitan areas and states, the prior standardized test scores of students transferring to charters differed slightly from those of their feeder school peers, and the direction of the difference varied across locations. In two locations (Chicago and Philadelphia), charter transfer students had higher test scores than their feeder school peers. In four locations (Ohio, Texas, Denver and San Diego), their scores were slightly lower than those of their feeder school peers. And in Milwaukee, there was virtually no difference in test scores between charter movers and their feeder school peers.23
The same study also found that Hispanic and African-American students who transferred to a charter school had lower prior standardized test scores than their same-race elementary school peers in the majority of sites. White transfer students, who were a small minority in most locations, tended to have higher prior achievement scores than their elementary school peers.24
Studying middle schools run by the nonprofit charter management organization Knowledge Is Power Program (KIPP), Mathematica found that in 2010–11, KIPP students tended to have slightly lower math and reading scores in elementary school than students in the same elementary feeder schools who did not apply to KIPP.25 However, students who transferred into KIPP middle schools in higher grades tended to have higher prior achievement levels than students already enrolled at KIPP.26
Finally, researchers who examined Florida students’ standardized test scores from 2000 to 2005 found that the better students were performing at traditional public schools, the less likely they were to transfer into charter schools.27
Critics have repeatedly argued that charter schools may be counseling out or otherwise “pushing out” lower–performing students in an effort to keep their average student performance high, enhance their reputations as high-performing schools or save money.
However, there is very little empirical research on the topic. One exception is a recent study by Ron Zimmer and Cassandra Guarino of student data from 2006–07 in an anonymous large urban school district. They found no significant differences in the prior performance of students who had left charter schools compared with those who had left traditional public schools.28
Much more research and data are needed on this question, especially studies that are able to estimate both national trends and the extent to which charter schools and management organizations may vary in their approaches to the needs of the lowest-performing students.