Skip to main content
v2026.11,986 entries · CC-BY 4.0

Common Core of Data (CCD): NCES School and District Dataset

The NCES Common Core of Data explained: what its nonfiscal and fiscal files hold, how school IDs merge, and what the CCD cannot tell researchers.

Written and maintained by CASRAI Editorial Board

Published

On this page

The Common Core of Data (CCD) is the U.S. Department of Education’s primary database on public elementary and secondary education, run by the National Center for Education Statistics (NCES). It is an annual, national database covering all public elementary and secondary schools and school districts, with roughly 100,000 schools and 18,000 districts. It is a dataset, not an academic standard: it has no connection to the Common Core State Standards for mathematics and English.

What is the Common Core of Data (CCD)?

NCES describes the CCD as a comprehensive, annual, national database of all public elementary and secondary schools and school districts. State education agencies (SEAs) report their schools and districts to the federal government through the EDFacts system, and NCES publishes the results as public-use files and through lookup tools. The CCD offers a School Locator, a District Locator, a Data Query Tool, a CCD Data File Tool, a Finance Data Visualization, a Mapping Tool and a Peer Search. Penn Libraries’ U.S. education statistics guide summarises the same scope: public schools and districts, with student population by grade, race and gender, classroom teacher FTE, the student-teacher ratio, and free and reduced-price lunch eligibility.

Researchers use the CCD as a sampling frame, a source of school and district context variables, and a way to describe the population of public schools in a study area. It is part of the wider set of federal education data programmes described in the CASRAI guide to the National Center for Education Statistics.

Common Core of Data versus the Common Core State Standards

The two share a word and nothing else. The Common Core State Standards are K-12 learning expectations. The Common Core of Data is a statistical collection about schools: who they are, where they are, how many students and staff they have, and how they are funded. If a source mentions CCD variables such as enrolment or free lunch eligibility, it means the NCES dataset. Searching with “NCES” or “CCD” in the query avoids the confusion.

What does the CCD contain?

The CCD has two components, nonfiscal and fiscal. The data table below this guide lays out each collection, the level at which it is reported and what it holds. In short, the nonfiscal CCD provides directory data, student membership counts by grade, race/ethnicity and sex, teacher full-time-equivalent (FTE) counts, staff FTE counts by professional category, counts of children with disabilities and English learners at the LEA level, and counts of students eligible for free and reduced-price lunches at the school level. The fiscal CCD provides revenues from local, state and federal sources, expenditures by function and object, average daily attendance at the SEA level, and student membership at the LEA and SEA levels.

Two units of observation matter for study design. A school is the unit for directory records, membership, teacher FTE and free and reduced-price lunch counts. A local education agency (LEA, or district) is the unit for disability and English-learner counts and for finance. A question about school-level Title I funding cannot be answered from the CCD: Title I funds are allocated to states and then to LEAs, LEA allocations come from the CCD fiscal collection, and school-level Title I allocations are not available. School-level Title I status is available in ElSi or the CCD school file.

How do NCES IDs work, and how do you merge CCD files?

NCES assigns a unique ID to each school and district when the SEA first reports it through EDFacts. The file layouts define the school identifier NCESSCH as a 7-digit agency ID (LEAID) plus a 5-digit school ID (SCHNO), 12 digits in all. The first two digits of the LEAID are the state FIPS code, and SCHNO is unique only within an LEA, so a school is identified by combining the two.

Identifier What it identifies Structure stated in the NCES file layout
LEAID Local education agency (district) 7-digit NCES agency ID; first two digits are the state FIPS code
SCHNO School within an LEA 5-digit NCES school ID; unique within an LEA only
NCESSCH Public school, across the whole file LEAID + SCHNO, 12 digits

The practical consequence is that these IDs are codes, not numbers. NCES instructs users to keep leading zeros in SEA, LEA and school IDs, and the EDFacts Read Me explains how to avoid dropping them when merging files. If a spreadsheet or import routine reads NCESSCH as a number, a school in a state with a leading-zero FIPS code will fail to match. Read ID columns as text in every tool. The identity of a school over time is a separate question: the CCD reference library lists longitudinal ID crosswalks, and the CASRAI term identifier crosswalk defines what such a table does.

What is the difference between preliminary and provisional CCD data?

A CCD file’s version label tells you how far its quality review has gone, and which one you cite matters for reproducibility.

Label Status What NCES says
0a, 0b, 0c … Preliminary Has not completed full data quality follow-up; may contain unresolved errors
1a, 1b, 1c … Provisional Reviewed and edited with state agencies; considered final

Either version can have several release rounds, so a 0b file supersedes 0a, and a 1a file supersedes any preliminary release for that year. The fiscal CCD uses fiscal years rather than school years: the fiscal year is the 12-month period of the annual operating budget, so FY18 corresponds to the 2017-18 school year. Aligning a finance file to a membership file therefore needs an explicit year mapping.

How do you work with CCD files in R, Stata or ElSi?

NCES warns that the raw files can be tedious to work with and that some, especially student membership files, exceed 1 GB. The membership files are long files, with separate records for enrolment by grade, sex and race, so they exceed Excel’s row limit and NCES says they must be processed in SAS, SPSS, Stata or R. For subtotals only, the ElSi table generator, which gives access to all years, levels and components, is the simpler route and needs no programming.

The key habit in code is to read identifiers as character. In R, readr::read_csv("membership.csv", col_types = readr::cols(.default = "c")) reads every column as text, after which counts can be converted with as.numeric(). In Stata, import delimited using "membership.csv", stringcols(_all) does the same, and destring converts the count columns afterwards. File names here are placeholders for the file you download.

Because the membership files are long, collapse before merging. Sum the counts to one row per school (or per school and grade) first, then join to the directory file on NCESSCH. Joining a long file to a wide one without collapsing multiplies the directory rows and double-counts anything summed afterwards.

What can and cannot the CCD tell you?

The CCD is a census of reporting public schools, not a sample, which removes sampling error but leaves definitional and reporting differences between states. NCES lists several limits directly:

  • Public schools only. NCES locators give IDs only for public schools and districts; private schools are covered by the separate Private School Universe Survey.
  • Not a school poverty measure. The CCD collects free and reduced-price lunch eligibility counts at the school level, but since SY 2016-17 states may report students certified through applications, directly certified students, or a combination, and some states report only one type, so counts are not comparable across states without checking how each reported.
  • No rankings. NCES states: “There are no accepted indicators or rankings of schools in this country.”
  • No manual edits to names and addresses. School names and addresses cannot be changed between publications; corrections go through the SEA.
  • New schools lag. A school that opened in a later year may not appear in the locator until its data are published.

For tracking identity over time, geographic data for schools and LEAs have been published as separate files since SY 2014-15, with congressional district codes available in ElSi and those files.

How do you cite and report CCD data?

State the file, the school year or fiscal year, the level (school, LEA or SEA) and the version label (preliminary 0a or provisional 1a, for example), and give the NCES page you downloaded it from with the access date. A methods sentence might read: “School enrolment and free and reduced-price lunch counts come from the NCES Common Core of Data, school-level nonfiscal files, provisional version 1a, merged on the 12-digit NCESSCH.” Naming the version lets a reader reproduce your numbers after NCES issues a later release. A structured data codebook, such as the DDI Codebook format, can carry the same information alongside your analysis files, and the data citation principle explains why datasets deserve a citation of their own. Student-level analyses that link CCD school data to individual records bring in privacy rules such as FERPA, though CCD itself is aggregate school and district data.

Sources

Data: Components of the NCES Common Core of Data (nonfiscal and fiscal)

Compilation of the CCD nonfiscal and fiscal components, the reporting level of each collection and its content, taken from the NCES CCD Data Files page.

Period covered: annual releases. Coverage: United States.

Components of the NCES Common Core of Data (nonfiscal and fiscal)
ComponentCollectionReporting levelContent
NonfiscalDirectorySchools, LEAs, SEAsDirectory data
NonfiscalStudent membershipSchools, LEAs, SEAsCounts by grade, race/ethnicity and sex
NonfiscalTeacher FTESchools, LEAs, SEAsTeacher full-time-equivalent counts
NonfiscalStaff FTELEAs, SEAsCounts by professional category
NonfiscalChildren with disabilities and English learnersLEACounts of children with disabilities and of English learners
NonfiscalFree and reduced-price lunch eligibilitySchoolCounts of eligible students
FiscalRevenuesNot stated on the data files pageLocal, state and federal sources
FiscalExpendituresNot stated on the data files pageBy function and object, including general administration, school administration, operations and maintenance, and facilities acquisition and construction
FiscalAverage daily attendanceSEAAverage daily attendance
FiscalStudent membershipLEA, SEAStudent membership counts

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Common Core of Data (CCD): NCES School and District Dataset

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →