Skip to: Content
Skip to: Section Navigation
Skip to: Site Navigation
Skip to: Search

Internet Use in Libraries

ALA Library Fact Sheet Number 26

How extensive is Internet Access?

Almost all academic, public, and school libraries in the U.S. are connected to the Internet, both for staff use and for public access. Here are the most recent national statistics for each type of library:


94.6% of libraries of postsecondary degree-granting institutions offered access to the Internet as of fall 1998. Access is almost universal in libraries of institutions with enrollments of over 1,500 (98%), but not always available in smaller institutions (91%).
(Source: Table 12.B in Academic Libraries 1998. Available at:


98.7% of public libraries are connected to the Internet and 95.3% of outlets provide public access to the Internet as of Spring 2002. Access is almost universally available in urban public library outlets, but not always available in suburban outlets (95.3%) or rural outlets (93.7%).
(Source: Table 4 in Public Libraries and the Internet 2002: Internet Connectivity and Networked Services. Available at:


90.1% of public school library media centers had an Internet connection as of the 1999-2000 school year. The level of connectivity is higher in secondary schools (96.9%) than in elementary schools (87.9%).
(Source: Schools and Staffing Survey, 1999-2000: Overview of the Data for Public, Private, Public Charter, and Bureau of Indian Affairs Elementary and Secondary Schools. Available at:

Extent of filtering in public libraries - Before June 23, 2003

The Supreme Court decision in United States v. American Library Association, No. 02-361 (June 23, 2003), "the CIPA decision," means that public libraries that accept federal dollars must install filters on all computers. But long before that date, some public libraries had installed filters on some or all of their computers. How many? John Bertot and Charles McClure of the Florida State University, Information Use Management & Policy Institute gathered data on that topic in the spring of 2002, through a grant from the U.S. Institute of Museum and Library Services (IMLS). Their report, Public Libraries and the Internet 2002: Internet Connectivity and Networked Services, was the 6th in a series of Internet connectivity studies conducted by this research team.

The study used a stratified sample drawn from the 16,004 public library outlets that could be geo-coded and sampled in terms of their level of poverty (less than 20%, 20%-40%, and greater than 40%) and metropolitan status (urban, suburban, and rural). Selected outlets received a brief questionnaire covering connectivity, number of terminals, access to subscription databases, and use of filters. Regarding filters for content (Table 10), the results were as follows:

None 52.1%
On Some Computers 17.5%
On All Computers 24.4%

Outlets in urban areas were most likely to have filters on all computers and outlets in rural areas were most likely to have no filters.

Filtering in schools and libraries - After June 23, 2003

For information on CIPA, consult the pages at:

As noted above, the Supreme Court decision in means that schools and public libraries that accept federal dollars must install filters on all computers.

The application of filters brings together several managerial strands: software and hardware selection decisions similar to those made for other aspects of library automation; managerial cost analysis, including cost-benefit analysis of the cost of applying filters against the benefit of receiving federal monies; and selection decisions to determine which sites to filter. The rest of this fact sheet will provide guidance on the software and hardware selection and cost issues.

Selecting a filter

In the September/October 2003 issue of Public Libraries, the official journal of the Public Library Association (PLA), Sara Weissman listed several factors to consider:

  • Will you be installing filters on a server or on individual PCs?
  • What are the terms of licensing?
  • What is the availability for installation: Online? CD-ROM?
  • How easy is the product to install and update?
  • Does the vendor publish a block list? How updated and how often?
  • Can you adjust or amend the block list? Are there levels of blocking from which you can select?
  • Does the product track sites visited? Is any log consistent with your patron privacy policy?
  • Does the vendor report to you the sites that your patrons visited? Does the vendor sell that information to third parties?
  • How easy is the filter to turn on and off?
  • Does the filter work with any security you've already installed on your machines?

And she recommends asking for names of other libraries using the product and checking on their experiences.
(Sara Weissman, "Filters: A Checklist for Product Selection," Public Libraries, September/October 2003, p. 279 © ALA. Used by permission.)

A more extensive online discussion, also from PLA, appears in PLA Tech Notes: Plain Facts About Internet Filtering Software, written by Karen G. Schneider.

The ALA E-rate Task Force has prepared the Adobe Acrobat Reader document, Sample Requests for Information Questions, that libraries may find useful to include in such requests that they send to vendors of filters or other technology designed to meet CIPA requirements.

A source of names of products used in libraries is the Children's Services Survey Section in the Public Library Data Service (PLDS) Statistical Report 2003. The survey for that report requested the name of the filtering software used by the public library on the computers available to children (if applicable), and these are indicated in the report.

The issues of providing filtered access, particularly in the school setting, are discussed in depth in "Safeguarding the Wired Schoolhouse: School District Options for Providing Access to Appropriate Internet Content" from the Consortium for School Networking (CoSN); access this statement at the CoSN web site at:

For a broader discussion on implementing technology in your library, see the ALA Library Fact Sheet 21 - Automating Libraries: A Selected Annotated Bibliography at which covers the automation basics of identifying system options, developing specifications, and working with vendors.

Filtering Costs

Among the costs to be considered are the license fee to cover the number of computers in your system, additional hardware required to run the filtering software (if any), training for the IT staff for installing the software, training for the library staff in its use (including knowing how to block and unblock sites), staff time to review, evaluate, and select the software; staff time to selected, implement and administer the filter; some of these are ongoing costs and others are the one-time start up costs.

Actual Cost of Filtering: A Few Examples

In early October, ALA identified a small number of school districts and public libraries who were known to use filters and who we expected would respond quickly. All filtered at the server level. We asked them three questions:
  1. How many computers do you filter now?
  2. What were your one-time costs for hardware (e.g. servers, peripherals)?
  3. What were your ongoing costs in the most recently completed fiscal year for
    1. License
    2. Staff
      (Estimate number of hours spent per year to administer filters and multiply that figure by staff member costs per hour including salary and benefits. Consider automation staff, collection management and admin staff.)

Some responded to all three questions, some were able to answer only two. Using that information, we calculated one-time costs per filtered terminal and ongoing costs per filtered terminal.

School responses came from five different states. The respondents who gave one-time costs reported a range of from a low of $0.20 per filtered terminal to a high of $7.33 per filtered terminal. The school reporting the lowest one-time cost had the highest number of computer (50,000) whereas the school reporting the highest one-time cost had the smallest number of computers (2,000). The respondents who gave us ongoing costs reported a range of from $1.30 per filtered terminal to $5.00 per filtered terminal.

Public library response comes from five different states. For those who reported one-time costs the range was from $6.20 to $25.00. Ongoing costs ranged from $10.88 to $33.94. Overall, the public libraries we surveyed are filtering fewer terminals -- from 90 to 845, as opposed to 2,000 to 50,000 -- meaning that costs were spread over fewer units.

For initial guidance on determining costs in your library, the E-rate Task Force has prepared several Excel spreadsheets to help library administrators capture, understand and compare the overall costs of filtering. These spreadsheets are in a beta test form and are available at the ALA CIPA web site, at:

For more information

Nancy Kranich, chair of the ALA Intellectual Freedom Committee, has prepared a bibliography of Useful Resources on CIPA, at:

The Washington State Library prepared a set of slides for its series of CIPA Workshops in 2003, which covered CIPA compliance, technology planning, filter selection and evaluation, and cost evaluation factors. There is also a bibliography of web-based resources. Access the slides online, at:

November 2003

For more information on this or other fact sheets, contact the ALA Library Reference Desk by telephone: 800-545-2433, extension 2153; fax: 312-280-3255; e-mail:; or regular mail: ALA Library, American Library Association, 50 East Huron Street, Chicago, IL 60611-2795.

Send suggestions for future presentations of this fact sheet to