While doing some research for a related project, I came across an interesting Pew Research Center Report, Library Services in the Digital Age (Zickuhr, Rainie, & Purcell, 2013). A particular finding that stuck out was the race/ethnicity differences in the attitudes towards library services and resources. In particular, Black or Hispanic individuals were more likely than White individuals to say that library resources/services were important to them and their community. The report also provides information on a host of potential new resources/services that Black and Hispanic individuals reported that they would more likely use.
Until recently, I have had limited experience working with large scale survey data (I’ve mostly worked with corpus and behavioral data from my graduate work or with data scraped from the web for my side projects). I thought pursuing this topic may be an interesting exercise. (Disclaimer: I, of course, am a psychologist in training so I apologize ahead of time to those with an expertise in library science or related fields in case what about I’m about to present is not particularly novel or a simplification! I do find this topic to be interesting and would love to hear from people with actual expertise in this area!). Given the difference in attitudes towards libraries among race/ethnicities, I was curious to see if there were differences found in libraries’ actual services/resources depending on the surrounding communities’ demographic makeup. Similarly, I was curious as to whether differences could be found in how different groups used libraries, not just based on what they report (as is the data discussed in the Pew report) but rather based on the reported activity of the libraries (e.g., number of annual visits, total attendance to programs, etc.).
To get at these questions, I looked at the Public Library Survey data that was collected in 2012 (this was the same year from which the data discussed in the Pew report was collected). The Public Library Survey dataset is a public dataset made available by the Institute of Museum and Library Services. The annual survey is collected across more than 9000 libraries across the United States and includes a host of information for each library such as its address, hours, total annual number of visits, total number of registered users, etc. Combined with some data that I gathered from the US 2010 Decennial Census, I was able able to get information regarding the general demographic makeup of the surrounding community of a library. For now, I focused on libraries located in cities. (In the future, I hope to look at libraries in different locales, such as rural areas, suburbs, etc., to determine whether the patterns found within city libraries are reflected in these other locales as well).
A library’s “community” here was defined as any area that fell within a 2 mile radius of a library. This is not to say that libraries only service the population that is located within a 2 mile radius. However, a cut-off was needed and 2 miles did not appear to be completely unreasonable given that in cities, libraries are typically more clustered in comparison to libraries located in other locales (e.g., rural areas). (I had initially hoped to use the mean/median distance from one library to another but there was just too much variability and I didn’t think the mean/median ended up being a useful marker.) Suggestions for what may be a better cut-off is welcomed! After determining the race/ethnicity makeup of the community (e.g., proportion who identified as Hispanic, who identified as Black, Non-Hispanic, etc.), I relabeled the overall community based on the majority race/ethnicity group. For more information on the different data sources I used and the analyses I performed, you can visit my github page for this project.
I should mention that these analyses I was able to perform was constrained by the available measures on the Public Library Survey and as such, they are not meant to be a one-to-one comparison to the results reported in the Pew report. For example, the Public Library Survey has limited information regarding the programs that involve introducing members to new technology (an item discussed in the report) so my analyses will not be able to speak to that. However, I think exploring some basic demographic differences would still be interesting.
Do the total number of annual visits and registered users differ by the demographic makeup of a library’s community?
We see differences in how often libraries are used based on the race/ethnicity of the surrounding community. In particular, libraries located in majority White communities typically have the highest number of visits (this is after controlling for the size of the population of legal service for each library). This is followed by libraries located in majority Black communities, with libraries located in majority Hispanic communities having the least number of visits out of the three. Similarly, libraries located in majority White communities again have the highest number of registered used. However, Black and Hispanic majority communities don’t differ in the number of registered users.
Does the library’s collection vary depending on the demographic of a library’s community?
To take into account different sizes of individual libraries, I looked at the size of different collection types (e.g., print material, Ebooks, etc.) per capita (i.e., population of legal service area). We again see demographic differences. Regardless of the material type (e.g., books or audio) and format (i.e., physical or digital), communities with majority White individuals have a larger collection compared to communities with majority Black and with majority Hispanic members. The latter two communities are comparable in their digital collection size; libraries in communities with majority Black members, however, have more physical material (i.e., print, audio, and video).
Do the library services used by members of a community differ depending on its demographics?
I looked at four services that had available data on the Public Library Survey: (1) total annual circulation, (2) total audience at programs, (3) total reference (e.g., information and referral consultations) transactions, and (4) total use of public internet computers.
In terms of borrowing material, communities with majority White individuals had more circulation transactions than the other two communities. In contrast, it was communities with majority Black members that had the most reference transactions. Majority Black and majority Hispanic communities also used computers for internet sessions more than communities with majority White members. Finally, demographic groups did not differ in how often they attended library programs.
Going back to our initial broader questions, we do indeed find demographic differences both in terms of the resources available in libraries as well as the library services being used by the surrounding community. Moreover, to some extent, these results are consistent with the findings reported in the Pew report (Zickuhr et al., 2013):
“Compared to whites, African-Americans and Hispanics are more likely to say libraries are important to them and their families, to say libraries are important to their communities, to access the internet at the library (and feel internet access is a very important service libraries provide), to use library internet access to hunt/apply for jobs, and to visit libraries just to sit and read or study.
For almost all of the library resources we asked about, African-Americans and Hispanics are significantly more likely than whites to consider them “very important” to the community. That includes: reference librarians, free access to computers/internet, quiet study spaces, research resources, jobs and careers resources, free events, and free meeting spaces.”
The lower rate of visits and registered users of libraries located in majority Hispanic and Black communities would seem to contrast the finding that Black and Hispanic individuals were more likely to report that libraries play an important role in the community. However, if we look at the resources that these Black majority and Hispanic majority communities used, we find that they do indeed make more use of the reference librarians and computers for internet access, consistent with the findings reported here. Thus, maybe “importance” may be better measured using something besides pure visit and user number metrics. Another interesting point is that with the interesting digital age, libraries located in Black majority and Hispanic majority communities still lag behind in their digital content compared to libraries located in White majority communities. To what extent this gap closes in the next few years will be interesting to explore. Additionally, as I had mentioned, these findings are constrained to libraries located in cities, with surrounding community being defined as a 2-mile radius from a library. Follow-up work is needed to determine whether these patterns are found in libraries located in other locale and to what extent increasing/decreasing our community size impacts these patterns.