Prison library statistics have received much less public attention than public library statistics. State and federal prisons about 1977 had 10.2 books per adult and circulated during the year on average 17 books per prisoner. Public libraries in 1977 held 2.3 books per adult served and circulated yearly 5.2 books per adult served. Compared to public libraries, prison libraries had higher book holdings per adult and greater book circulation per adult.
The data source for prison libraries, a 1977 survey, falls far short of encompassing the universe of state prison libraries. That survey does not include any prison libraries in 13 states. The total number of prisoners in reporting prisons amounts to only 49% of prisoners in state prisons at year-end, 1976. In the 1974 census of state correctional facilities, 83% of responding facilities reported having a library, and these facilities encompassed an estimated 93% of the prison population. Hence the 1977 survey totals need to be adjusted to estimate figures for the universe of state prison libraries.
To estimate book volumes in prison libraries nationally in 1977, we scaled reported book volumes by an estimated total number of prisoners, adjusted for prison type and library coverage. The prison type most likely to have a prison library is a closed prison. In the 1974 state prison census, closed prisons held 63% of all prisoners. We used this share, along with a figure for the total 1977 state prison population, to estimate the number of prisoners in closed prisons in 1977. We then scaled reported volumes in prison libraries in closed prisons by estimated total prisoners in closed prisons. We used a corresponding calculation to scale reported volumes in prison libraries other than in closed prisons. The estimated U.S. state prison library book holdings in 1977 is 2.8 million volumes.
Other estimates about 1970 for total volumes in state prison libraries have weaker documentary and weaker analytical support. As part of the American Library Association’s National Inventory of Library Needs, the American Correctional Association and the Association of Hospital and Institution Libraries about 1964 surveyed state and federal correctional institutions for persons 16 years and over. Survey responses were received from 150 institutions out 294 surveyed, but individual institutional responses are not available. Total volumes reported in libraries of state correctional institutions was 1.0 million. This total includes no volumes for institutions in eight states. Figures for total adult inmate population, which were then used to calculate total volumes needed and the shortfall in volumes, only correspond to the total adult inmate population of reporting prisons when replies were received from all prisons in the given state. That’s a major analytical weakness. Given available data, the best method for estimating total volumes for the universe of prison libraries is scaling total reported volumes by the survey response rate. Such scaling implies 2.0 million volumes in state prison libraries about 1963.
A study of correctional libraries in 1970 also presents major problems for estimating aggregate volumes in state prison libraries. This study is based on a survey of state institutional library consultants. State institutional library consultants depended for their statistics on responses from individual state correctional institutions. Neither the responses of individual state correctional institutions, nor those of the state institutional library consultants, are available. State library consultants in 42 out of 50 states responded to the survey. However, only 37 out of 42 responses provided data on volumes in state prisons libraries. The extent of completeness in reporting of individual institutions to the 37 states that provided volume data is not known. The 42 states that provide some response apparently encompassed 73% of state prisoners. Given available data, the best method for estimating volumes in the universe of prison libraries is scaling volumes from 37 to 42 reporting states, and then scaling by a 73% state prisoner population coverage ratio. This procedure does not account for incomplete reporting at the state level. It implies 1.5 million volumes in state prison libraries in 1970.
Additional data exist on book holdings in federal prison libraries. Twenty federal correctional institutions responded to 1972 survey of federal prison libraries. Scaling the reported volume of books by share of federal prisoners in reporting prisons (45%) implies 319,000 volumes in federal prison libraries in 1972. The National Inventory of Library Needs found 265,582 volumes in federal prison libraries about 1964. Some qualitative evidence indicates that federal prison libraries did not expand their holdings between 1972 and 1977.^ We take the 1972 estimate of volumes in federal prison libraries as an estimate for holdings in 1977.
The share of prisoners using prison libraries is much more difficult to estimate than the number of prisoners in prisons. From 1945 to 1992, various sources give prison library user shares typically from 50 to 70%. A statistical survey of prison libraries, with responses from 79 mainly state prisons, reported in 1945 the “percentage of readers to population”:
The average from all reports is a trifle over 54 per cent, which is 10 percent under the figure reported last year. That means that according to the survey, only half of all prisoners are regular readers. Actually, the figures must surely be much higher.^
A survey of federal prisons about 1950 survey indicated that “75 percent of all federal prisoners are making use of their library privileges.”^ Reporting on responses from 120 state prisons and 28 federal prisons, an educational official at a federal penitentiary stated in 1950 that in prisons for adults with population less 500 prisoners, “between 70 and 75 percent of the inmates use the libraries.” For prisons with population greater than 500 prisoners, “just 60 percent of the state and 80 percent of the federal inmates use their libraries.”^ We estimate the prisoner library user share to be 66% for 1977. That’s the same estimate for user share used for the summary estimates of circulation per prison library user c. 2000.
Relevant dataset: prison libraries’ book holdings and circulation, 1940-1980