How do we identify and count charter schools?

The National Alliance for Public Charter Schools manages a database that serves as a repository for charter school data. The National Alliance collects data from a variety of sources including:

  • State data files that contain enrollment, directory, and performance information.
  • Federal datasets, such as Common Core of Data (CCD), the Civil Rights Data Collection (CRDC), and EdFacts.
  • State and local partners and, in some cases, individual schools.

In addition to reviewing the quality of the data verifying the charter status, our team processes and categorizes the data based on the type of record.

How do we categorize schools?

No exact definition of what constitutes a “school” exists. Because federal and standardization can be problematic when accounting for schools in data systems and in maintaining consistent data across time. For those interested in gaining a better understanding of the education landscape, these inconsistencies pose a problem, particularly in the counting of charter schools. A single record of a school (typically represented as a row or unique identification number in each dataset) is not necessarily a single school. For example, the record may represent a program within a school or a campus associated with another school. These record types can vary between federal and state datasets and appear frequently among charter school records. That means comparing or combining federal and state datasets without properly understanding, and ideally rectifying, these discrepancies can lead to inaccurate counts of charter school openings, closures, and other important metrics. To address this, the National Alliance developed a “Modified Count” system.

Each database record receives a Modified Count code. The Modified Count code indicates whether the school is a campus, charter holder, program, or indeed a school. Accounting for and correcting these discrepancies when possible is essential for generating accurate counts of charter schools, openings and closures, and other important metrics. The National Alliance’s Modified Count code allows researchers to consistently classify records across data sources and school years.

Charter Record Type by Year

School Year Schools Campuses Holders School Share Campus Share Holder Share
2015-16 6532 527 8 92.4% 7.5% 0.1%
2016-17 6675 554 10 92.2% 7.7% 0.1%
2017-18 6776 575 10 92.1% 7.8% 0.1%
2018-19 7106 477 10 93.6% 6.3% 0.1%
2019-20 7241 458 10 93.9% 5.9% 0.1%
2020-21 7371 451 10 94.1% 5.8% 0.1%
2021-22 7547 449 9 94.3% 5.6% 0.1%

Some states appear disproportionately affected by campuses and holder records (i.e., nonschool records). Nonschool records are administrative records that typically hold enrollment for two or more schools. States with some of the largest charter school sectors, such as Washington, D.C., Texas, and Florida, also have some of the highest proportions of nonschool records. For more information on this issue, please see our 2019 Modified Count Report. This method of coding is especially important when looking at new and closed schools. A record might appear as a new holder, but the same number of schools continue to operate. This would not count as a new school. Similar, only new records that exist as either additional campuses to previous schools or new schools altogether count towards new schools.

States With More than 5% of Records as Non-Schools

State Holders Campuses Share of Non Schools
AL 0 2 16.7%
AR 1 11 13.0%
AZ 1 39 6.8%
CO 0 17 6.2%
DC 1 21 16.5%
FL 1 78 11.2%
MI 1 22 6.2%
MN 0 34 12.1%
MO 0 11 14.1%
MS 0 1 14.3%
NH 0 5 13.9%
NV 0 7 7.4%
OH 0 17 5.2%
OK 0 14 20.6%
RI 0 3 7.5%
TN 0 12 10.3%
TX 0 97 9.6%
WY 0 1 20.0%

How do we account for changing school identification codes?

State and federal National Center for Education Statistics (NCES) school ID codes sometimes change from year to year. The National Alliance Data & Research Team developed an identification system to address this, where we merge data into our Database of Charter Schools. The system establishes a unique NAPCS ID that allows the team to identify schools across time, regardless of whether state or federal IDs have changed.

State and federal IDs are typically embedded with information—for example, a code to identify a school’s district or county. If underlying information relating to the code changes, such as when the school is redistricted, the ID will change as well. In this situation, the school remains operational, but the school’s original ID may be eliminated, making it hard to determine if the school is still open or to track the school properly over time. In other words, the creation or elimination of a state or federal school ID is not necessarily connected to a school’s operational status. Accounting for these sorts of changes is important for accurately counting school openings and closings, especially for charter schools. Changes in IDs without careful attention might incorrectly look like the closing of schools and opening of new schools. A previous analysis found that between the 2005-06 and 2019-20 school years, federal NCES IDs for charter schools changed more than 500 times. However, NCES continues to update their IDs and presents the latest available ID, making this somewhat difficult to track.

The reasons for changing IDs vary depending on state context. For example, in some cases, an authorizer change may cause an NCES ID change. In the case of Louisiana, numerous schools received changes to their NCES ID when their authorization moved from the Recovery School District (RSD) to the Orleans Parish School Board. Conversions, mergers, and expansions are other factors that can lead to an NCES ID change.

How is demographic data collected?

Demographic data for student racial/ethnic backgrounds were collected using a combination of CCD and state-level datasets. We aggregated demographic data for “Other/Two or More Races,” “Native American/Alaskan Native,” and “Pacific Islander/Hawaiian Native” classifications into a single category labeled “Other,” along with any missing demographic enrollment, for the purposes of this report. As individual groups, these populations constituted very low percentages of the overall population in both charter and district schools with a relatively steady trend. However, “Other/Two or More Races” alone reached greater than 4% of the population for the first time in 2010-11.

Demographic data for students with disabilities comes from the U.S. Department of Education’s Civil Rights Data Collection. We calculated free and reduced price lunch (FRPL) data using only schools that reported data. Schools with no FRPL data were removed from the calculations. In 1998-99, the Free Lunch Eligible Status shifted to Free and Reduced Priced Lunch, or the more standard FRPL Eligible.

How are missing data addressed?

The National Alliance prides itself in its ability to collect and assemble the most up-to-date and thorough data on charter schools. However, some gaps in data inevitably occur. We calculate all statistics throughout the digest using the best available data and exclude any records for which no data were available.

How Do We Distinguish School Districts?

Because charter schools often serve as their own districts, we assign all charter schools a geographic school district in order to find school and enrollment share. We acquire addresses for every public charter school and district school from Common Core Data (CCD), as well as our own collection process. We then use tools from Geocodio to find latitude and longitude values for each school and use ArcGIS software to plot those values on public school district shape files. This allows the National Alliance’s Data and Research Team to see which geographic school district a given charter school would fall into if it were coded as part of that district. Locale data is derived from CCD.

The National Alliance eliminates known virtual schools* in enrollment share and locale analyses. Most virtual schools are flagged by CCD, but the National Alliance’s Data and Research Team also conducts a manual check for cyber and hybrid schools. Further, the team checks for schools with student populations far exceeding what is expected based on location information, which often indicates virtual enrollment.

*Note: Data in the 2023 edition of the Charter School Data Digest includes data from the 2005-06 through 2021-22 school years. The term “virtual schools” in this report does not apply to schools that have adopted virtual, remote, hybrid, or other social distancing practices in response to the COVID-19 pandemic unless this change was permanent.

Measuring School Finances

The National Alliance obtained data on school finances from the 2006-07 through 2019-20 National Public Education Financial Survey. The survey indicated whether a given district has charter schools. To isolate the financing of each sector, we only use data from districts that are exclusively charter schools or exclusively district schools. The Local Education Agency (LEA) districts that consisted of both charter and district schools were removed from the analysis because we had no way of knowing the allocations between sectors. Because of this, we do not have the complete universe of data for either charter or district schools. Still, we account for approximately 65 percent of charter school students, though we notably are missing data from New York and Florida, two states with some of the largest charter school sectors.

The unweighted average was used for each state and nationwide in this analysis. The formula for Per Pupil Revenue is defined as the sum of TOTALREV/sum of student enrollment. 740 LEA districts in the 28 selected states either have missing revenue data (TOTALREV = -1), revenue data not applicable (TOTALREV = -2), or no revenue (TOTALREV = 0), and they were removed from this analysis. The 28 states were selected in this analysis because they had lower rates of suppression/missing data compared to other states, and most of their districts were charter-only or district-only school districts.

About the Authors

Jamison White
Jamison White

Director, Data and Research

Before joining the National Alliance in 2017, Jamison worked as a financial and small-business consultant in Pittsburgh, Boston, and the greater New York area. Jamison studied at Carnegie Mellon University and Philipps-Universität Marburg, Germany. He is a part of a founding group for a classical charter school in Washington, DC. In his free time, Jamison researches school curricula, pedagogies, and charter school models.

Cynthia Xu
Yueting "Cynthia" Xu

Senior Manager, Data and Research

Yueting "Cynthia" worked as an ESL instructor and education consultant in Philadelphia prior to joining the research team at the National Alliance. During her undergraduate years at Sun Yat-sen University, she studied English language & literature and Economics. She received her master’s degree from the University of Pennsylvania with dual majors in ESL education and statistical measurement & research.