Survey notes

The survey and the unit of comparison

The survey counts administrative entities: the legally established library system, which may run one building or dozens of branches and bookmobiles. A system page is therefore about the whole organization, not a single branch. Each library reports its own twelve-month fiscal year; depending on the state that is July 2023 to June 2024, October 2023 to September 2024, calendar 2024 or a local variant. The page header shows each system's own period.

Which systems are profiled

The FY 2024 system file holds 9,249 records. We set aside the 44 records from the outlying areas, 171 entities that do not meet the federal definition of a public library, 4 systems closed for the year and one statewide service for blind and print-disabled readers. We then keep only systems whose service population, visits, physical checkouts, operating spending, print collection and outlet counts were reported by the library itself; IMLS fills gaps with estimates, and 429 systems had at least one of these core values estimated or missing. A handful more were left out for having no building, a service population under 100, zero visits or spending, or ratios outside any plausible range. That leaves 8,561 profiled systems.

Secondary items such as open hours, program attendance, computers, staff and the revenue breakdown are shown only when the library reported them. An estimated value never appears; where it would have been, the row or chart is simply absent. In all, 314 systems have their own page: every system serving 250,000 people or more, the largest system in each state, and systems serving 25,000 or more that hold at least four top or bottom placings. All others appear in the peer check and in state and size-class tables.

What "residents" means

Per-resident figures divide by the population of the library's legal service area, the area that funds it and that it is set up to serve, as reported by the state library agency. In Arizona, Colorado, Connecticut, Florida, Indiana, Iowa, Kentucky, Maine, Massachusetts, Minnesota, Mississippi, New Hampshire, New Jersey, New York and Vermont some service areas overlap (a county system and a city library can both count the same households), so a county system there can look thinner per resident than it is. State totals per resident on this site use the IMLS unduplicated population, which removes that double counting.

How each measure is calculated

Visits per resident	Annual visits ÷ service population.
Checkouts per resident	Physical items lent in the year, renewals included ÷ service population. Total circulation is not used: its definition changed in FY 2024 and a third of systems did not report it.
Spending per resident	Total operating expenditure (staff, collection, other operating costs; capital excluded) ÷ service population, in nominal dollars.
Print volumes per resident	Books and other print items held ÷ service population.
Program visits per 1,000 people	Attendance at live (synchronous) programs × 1,000 ÷ service population.
Computers per 10,000 people	Internet computers for public use × 10,000 ÷ service population.
Open hours per outlet a week	Annual public service hours of all outlets ÷ (central libraries + branches + bookmobiles) ÷ 52.
Staff per 10,000 people	Paid full-time-equivalent staff × 10,000 ÷ service population.
Local share of revenue	Operating revenue from local government ÷ total operating revenue.
Electronic share of checkouts	Electronic circulation ÷ (electronic + physical circulation), where both were reported.
Change since FY 2019	FY 2024 value ÷ FY 2019 value − 1, matched on the IMLS library identifier, shown only where the service population moved by less than 15%.

Size classes, ranks, medians and percentiles

Systems are compared within eleven population classes, from under 1,000 residents to a million or more, the same bands IMLS uses in its benchmarking tables. A rank such as "third among the 122 profiled systems serving 250,000 to 499,999 residents" counts only systems that reported that measure; tied positions are never presented as a rank. State ranks count profiled systems in that state. The median is the middle system of a class. A percentile is the share of the class with a lower value, with ties split evenly; on a system page the spine height shows it. The six peers in the comparison table are the profiled systems closest in service population anywhere in the country.

Limits worth keeping in mind

Libraries count visits differently: some use door counters for the full year, others multiply a sample week. Dollar figures are not adjusted for inflation or local costs. States define service areas in their own way, and Hawaii reports its whole public library service as one system. Comparisons describe what was reported; they do not rate the quality of a library or say how it should be funded.

Source and licence

Institute of Museum and Library Services, Public Libraries Survey, Fiscal Year 2024: data files and the Data File Documentation and User's Guide (June 2026); trend bases from the FY 2023 and FY 2019 files. IMLS states that its survey information "is in the public domain and may be used and reprinted without special permission", with citation required. All ranks, ratios and comparisons were computed for this site. The site and its operator are not connected with IMLS or any library.