© Copyright 2026 Mastersinlibraryscience.org.
A practical guide to building-level staffing data, FTE counts, and the definitions that make your numbers hold up.
In 2025, Michigan lawmakers weighed a mandate for a certified librarian in every public school. State fiscal analysts testified they could not price it, because no one could determine how many schools even have a library. Third-grade reading proficiency near 40 percent forced an emergency building-by-building census.
That gap is an information-science problem before it is an advocacy problem: unit of analysis, headcounts versus full-time equivalents, survey cadence, credential coding. Hallie Rich's Nov. 9, 2026 editor's note in School Library Journal (https://www.slj.com/story/We-Must-Rigorously-Track-School-Library-Access-Staffing-Editors-Note) makes that case directly.
The stakes are already documented: in Michigan, roughly one third of public school districts operate with no library staff, and the losses concentrate in high-poverty communities and the primary grades.
Which federal dataset can tell you whether a particular school building had a certified librarian on October 1?
None of them does, exactly. Two instruments cover parts of the question; knowing which part each one covers keeps a citation from becoming an overclaim.
The CCD is an annual NCES collection covering roughly 100,000 public schools and about 18,000 districts,1 split into nonfiscal and fiscal components. For 2025-26, states submitted in spring 2026 using an October 1, 2025 snapshot, and preliminary directory files appeared in July 2026.2 Release timing is uneven: recent nonfiscal files have arrived in under four months, while an earlier cycle took 20 months.3 Staffing files carry school and district counts, but the public documentation does not establish that library media specialists, librarians, and library aides are broken out as separate certified categories at each reporting level.4 Certification and licensure are not fields you can count on.
NTPS is planned every three school years, and it is the only federal instrument that captures a building-level view.5 The 2020-21 teacher questionnaire asked how many staff held full-time positions as library media center aides around October 1, 2020.5 That item drew an 82.8% response rate, and its documentation appeared in 2024.5 It counts heads, not full-time equivalents, and one sample year is not this fall.
Certification status. FTE rather than headcount. Whether a library exists in each building. Weekly access hours. Until those four are collected at the building level on a short cycle, the gap is a data design problem, not a complaint.
In 2025, Michigan legislators considered requiring a certified librarian in every public school. When fiscal analysts tried to price the mandate, they testified they could not: the state had no count of how many schools even have a library. Third-grade reading proficiency was reported at under 40 percent, and lawmakers responded by funding a complete building-by-building enumeration of more than 3,500 public schools rather than another sample. That census launched in September 2026 through Wayne State, and by late October about a third of invited schools had responded. Michigan still recommends certified librarians rather than requiring them. The mandate stalled; the state went looking for its baseline instead.
A district roster answers a different question than the one advocates think they are asking. One librarian reported on paper may actually be split across five buildings. Or a district's library staff may sit entirely at the high schools while elementary students go without. Aggregate to the district and every building appears covered. A district total that looks reasonable beside a state average can still conceal five schools with nothing in them. In 2026 reporting, fewer than 10 percent of Michigan schools reported a library staffed by a trained librarian, and another count put the share without a full-time certified librarian at 92 percent. About a third of districts operate with no library staff at all.
Measure at the building, then roll up. Never roll up from a district roster and assume even distribution across schools. Log shared and traveling librarians per building as fractions: 0.5 at one elementary, 0.2 at another, zero at a third. Those fractions should sum to no more than the total staffing your district reports. Partial coverage written down as partial is evidence. Partial coverage averaged into a district total is silence.
When nobody can say which buildings lack staff, cost estimates for a mandate become guesswork, and guesswork is where well-meaning bills quietly die. Building-level counts are not a reporting chore. They are the price tag.
The same person can appear in four different systems under four different labels, at four different levels of aggregation. This crosswalk lines up the terminology, the reporting unit, and the counting rule for the sources most often cited in school library advocacy, so that a district total is not read as a building figure and a headcount is not read as an FTE.
| Data Source | Term Used | Reporting Level | Certification Captured? | Headcount or FTE? | Question It Can Answer |
|---|---|---|---|---|---|
| NCES Common Core of Data, state-level record | Librarians | State | No. The cited glossary definition specifies no certification field. | FTE | How many FTE professional librarians and supervisors are reported statewide? |
| NCES Common Core of Data, district record | Librarians/Media Specialists | District | No certification field specified in the cited definition | Count of professional staff members and supervisors assigned library and media duties | How many professional librarians or media specialists does a district report in total? |
| NCES Common Core of Data, school universe record | Librarian; Media Specialist | School (public elementary/secondary universe classification) | No certification requirement stated in the cited definitions | Staff assigned specific duties and school time | Which staff select, catalog, circulate, and guide users, and which staff direct or supervise media centers? |
| NCES Common Core of Data, school universe record | Library and Media Support Staff (includes library aides) | School | No certification requirement stated | Included within the support-staff classification | How many aides and other support staff provide library or media services? |
| NCES National Teacher and Principal Survey | Library media center; library media specialist | School | Yes. Distinguishes a teacher who is state certified in the field of library media. | Headcount, counted per person rather than as FTE | Does the school operate an organized collection of print, audiovisual, and computer resources administered as a unit for students, teachers, and administrators? |
| State education agency staffing report (Kentucky example) | School media librarian | School | Yes. The position is described as requiring the appropriate certificate of legal qualifications under state law. | Not specified | Does the school employ a certified school media librarian to organize, equip, and manage media-library operations? |
| District HR job codes | Local titles mapped onto library media specialist, library aide, or media clerk | District personnel and payroll records | Sometimes, where the code is tied to a license or endorsement | Usually position-based and expressed in FTE on payroll | Who is on payroll in this building this year, and at what fraction of full time? |
| AASL professional terminology | School librarian, with teacher librarian and school library media specialist as alternates | PreK-12 professional role, not a reporting unit | Names the professional role; state licensure titles vary | Not a count | What is the official professional title for the person responsible for PreK-12 library and information services? |
The vocabulary of school library staffing is itself under revision, and the labels a district chooses shape what its numbers appear to prove.
AASL treats "school librarian" as the umbrella term for a professional who serves as leader, instructional partner, teacher, information specialist, and program administrator.1 States translate that into their own credential titles: teacher librarian in California, where the Teacher Librarian Services Credential requires a teaching credential first; library media specialist in Delaware, Maine, and Maryland3; school library media specialist under New Jersey's endorsement code 2855; and an ND or NQ endorsement in Michigan.2 Paraprofessionals and library aides sit outside all of it as noncertified support personnel, sometimes called library technicians or assistants.
A district can truthfully report that every building has library staff when that staff consists entirely of classified aides with no certification and no library coursework. Blended totals hide that difference, and blended totals are what budgets get approved on.
The rule for local counts is simple: certified and classified staff belong in separate columns, never merged into one "library staff" figure. The definitions are also contested. Michigan's HB 4922 proposed eliminating school library credentialing standards to ease hiring, which shows that the definition of who counts as a librarian is a policy lever, not a neutral label. The CCD carries no certification field, so local personnel records have to supply it.
EDFacts asks districts to report staff FTE to the nearest hundredth, so a library total may land as 59.20 rather than a whole number.1 That precision is the point. The working formula is weekly hours assigned to library duties divided by the full-time weekly standard the reporting system uses. Forty hours is the common denominator: Oregon and California both convert on a 2,080-hour year, South Carolina divides part-time hours by 40 and adds full-time employees, North Carolina divides total weekly hours by 40, and Massachusetts set its FY2026 basis at 40 hours per week rather than 35.2 A 37.5-hour standard can apply where a district adopts one, so always name the divisor.
Picture a building with one full-time librarian, one 20-hour library aide, and one librarian whose week is split across three schools. Headcount says three people. FTE says something else: 1.0 for the librarian, 0.5 for the aide (20 divided by 40), and 1.0 for the split librarian, allocated at roughly a third to each building. The district total is 2.5 FTE, and each school's report shows about 0.33 of a shared position rather than a phantom full-timer.
The National Teacher and Principal Survey counts heads. Five half-time librarians become five librarians in a summary table and 2.5 FTE in a budget conversation. That gap is exactly where an advocacy case weakens, because an administrator can answer the five with a single hire.
Publish heads and FTE side by side, state the hour standard you used, and follow the 2026 CRDC method for split assignments: calculate total district FTE, then multiply by each school's share of scheduled time.3 That keeps local figures benchmarkable against state and federal published totals.
State education agencies are the most reliable source for building-level library staffing, but the files are scattered and rarely share definitions. California offers the clearest current example of a state that publishes school-level data with certification detail.
Look for three things before you download: the reporting year, the staffing definition (certified librarian, emergency credential, teacher, paraprofessional), and whether certification status is its own field. That last variable is the one most state files omit. Oregon and Minnesota have also published school library reports in recent years, and Michigan's building-by-building library census answered a specific legislative question; confirm the current year and format on each agency site rather than assuming the file is maintained.
Many states do not release school-level library staffing files. If yours is one of them, do not wait for a statewide dataset. The local collection protocol in the next section gives you a defensible way to count buildings, staff, and access hours yourself.
A one-page brief built around a stated method beats a louder number that has none, every time a board votes. The method note is the part that survives cross-examination: a defensible count with a stated unit of analysis, a stated FTE standard, and a collection date holds up under a skeptical board member's questions where a bigger, vaguer figure does not.
Every brief sent upward should fit on a single sheet in this order:
Boards do not vote on FTE. Translate it. "Two elementary buildings have no certified librarian this year. Students there receive zero scheduled library minutes per week, against 90 minutes at the two buildings with coverage. Restoring 1.0 FTE across those buildings would serve roughly 780 students."
Keep certified and classified counts separate in every table that leaves your building. A paraprofessional running a library is real coverage, and combining the two invites a board to read a total that does not exist. Where one librarian serves several buildings, disclose the split building by building rather than reporting a district total.
State agencies can aggregate, audit, and publish what districts submit. They cannot retrofit building-level detail onto a federal instrument that was never built to carry it. Ask your state to collect building-level, certification-specific data in its own survey, and say so plainly in the brief.
The most useful number for local advocacy may not be a test score at all, but a visit count.
The best-known evidence comes from the Lance-era state studies. In Colorado, schools with endorsed librarians reported 2 to 3 percentage points fewer unsatisfactory reading results, described as a 15 to 25 percent proportional difference, and high schools with 60 or more library-media hours per week reported a 22.2 percent test-score difference. In Pennsylvania, roughly 8 percent more students reached the advanced level in reading where a full-time certified librarian was on staff.1 South Carolina's 2013 survey found elementary and middle schools averaged four weekly group visits, and that four or more visits were associated with writing strengths; high schools with 15 or more group visits were more likely to meet the HSAP benchmark.2 A 2026 California study linked any high-school library staffing to a higher first-semester university GPA, with a mean of 2.65 and a standard deviation of 1.17.3
None of this establishes that staffing caused the gains. These are observational studies, confounded by school demographics, funding, prior achievement, leadership, and which students choose to use the library. Report them as associations, every time.
Access frequency is the honest local metric. Track scheduled class visits per week, open library hours per week, and sign-in or circulation counts by grade and by building. Those numbers are yours to collect and defend.
Pair every outcome claim with a local access measure. Borrowed effect sizes collapse under cross-examination; a documented drop in elementary visits does not.
When Michigan lawmakers weighed a mandate for a certified librarian in every public school, state fiscal analysts told them they could not price it out, because no one could say how many schools even have a library. That is the practical cost of a missing baseline: a policy debate conducted in the dark.
A position cut only reads as a cut if the prior year's number exists to compare against. Record full-time equivalents by building every fall, before budget drafting begins. When a district proposes converting a librarian slot into two part-time aide roles, the documented drop from 1.0 FTE certified to 1.0 FTE classified is what turns an anecdote into a line item a board can act on.
Reconsideration debates hinge on questions staffing data can answer: who reviews requests, whether a certified librarian is on staff, and whether the library is staffed and open at all. A building where the library is unstaffed three days a week cannot run the same review process as one with a full-time librarian. Report that schedule alongside your collection policy.
Michigan opened its building-by-building census only after third-grade reading proficiency fell to nearly 40 percent, and a separate bill (HB 4922) proposed dropping credentialing standards to ease hiring. Both the unfunded mandate and the credentialing rollback moved faster because credible building-level counts did not exist.
Start the spreadsheet this year. Every future budget and challenge argument depends on a baseline someone collected before the fight.
