An attempt was made to identify the demographic characteristics that separate library use/non-use in Japan in the 2010s. This is a secondary analysis of the first SSP survey data. The following data were taken: gender, age, education, personal income, household income, homeownership, number of family members, marital status, presence of children under elementary school age, social class of origin, paternal education, maternal education, occupation, and city size of the area where the respondent lives. The relationship between these factors and frequency of library use was examined using ordinal probit regression analysis. The results showed that being female, having a high level of education, having a low individual annual income, having a high household income, owning a house, having young children, having a high paternal education level, and having a white-collar job were highly correlated with the frequency of library use. Further analysis of the results for the more-educated and less-educated groups showed that the correlation with gender and occupation weakened for the more-educated group, while the correlation with personal annual income and maternal education strengthened for the less-educated group. The implications of these characteristics for library use could be interpreted in relation to life stage, social class, reading ability, and opportunity cost.

　In this study, a survey was conducted using search results of “Catalog of Reference Books” and bibliographic data of the National Diet Library holdings with the aim to clarify the publication trends of reference books in Japan. Numbers of reference book publications calculated by referring to the “Catalog of Reference Books” from 1990 to 2019, and bibliographic data of reference books from 1950 to 2020 collected by referring to National Diet Library Subject Headings and National Diet Library Classification, were analyzed.
　The result showed that the numbers of publications of most reference books peaked from the 1980s to the 2000s and then continued to decline, earlier as a whole than general published books. It was also found that chronological tables, glossaries, encyclopedias, and bibliographies excluding those in the humanities showed a remarkable decrease, while bibliographies mainly in humanities, dictionaries in the natural sciences, and directories showed a gradual decrease or even an increase.