Public Library Density and Adolescent Reading Outcomes: Evidence from a Natural Experiment in Wisconsin
- Design
- Difference-in-differences, 2010–2024
- Treatment
- 2014 Wisconsin Library Construction Grant Program
- Sample
- 412 public school districts
- Outcome
- 8th-grade Forward Exam reading scale scores
- Identification
- Pre-program library-density rollout
Between 2014 and 2018, the State of Wisconsin disbursed approximately $74 million in capital grants for the construction or major expansion of public library facilities, with awards distributed across geographies on the basis of a pre-announced scoring rubric. We exploit the resulting variation in branch density to estimate the effect of public library access on adolescent reading achievement, using a difference-in-differences design with district and year fixed effects.
We find that the addition of one public library branch per 10,000 residents in a school district raised mean 8th-grade Forward Exam reading scale scores by 0.21 standard deviations (95% CI [0.13, 0.29]). Effects were concentrated in districts with below-median household income and were larger for students whose primary household language was not English. Our results are consistent with a model in which library access supplements rather than substitutes for school-based reading instruction.
1. Introduction
Despite a substantial literature on the educational effects of school-based interventions, the contribution of out-of-school institutions — and public libraries in particular — to adolescent learning outcomes remains comparatively underexamined. A small body of cross-sectional work has documented positive correlations between library access and reading achievement (Krashen, 2004; Constantino, 2005), but these designs cannot credibly identify causal effects given persistent selection of high-resource communities into both library investment and educational outcomes.
This paper exploits a 2014 Wisconsin state initiative that disbursed capital grants for public library construction across districts on the basis of a pre-announced rubric. The resulting rollout produced quasi-random variation in library density that is plausibly exogenous to district-level reading outcomes.
2. Data and identification
Library facility data were obtained from the Wisconsin Department of Public Instruction's Public Library Annual Report. Reading achievement was measured using district-mean scale scores on the 8th-grade Forward Exam (the state's standardized assessment, administered annually since 2015–16). We restrict the sample to 412 districts with continuous data across the 2010–2024 panel.
Our preferred specification is a two-way fixed effects difference-in-differences model with district and year fixed effects and time-varying controls for total per-pupil instructional spending, free/reduced-price lunch eligibility, and district urbanicity (RUCA). Standard errors are clustered at the district level.
3. Results
Adding one public library branch per 10,000 residents in a district raised mean 8th-grade Forward Exam scale scores by 0.21 standard deviations (Table 3, column 4). The estimate is stable across specifications and is robust to the inclusion of leading and lagged treatment indicators.
Heterogeneity analyses indicate that the effect was approximately 1.7 times larger in districts in the bottom half of the household-income distribution. We also find larger effects in districts in which more than 10% of students reported a non-English primary household language, consistent with prior work suggesting that library access disproportionately benefits households with constrained at-home access to age-appropriate English-language reading material.
4. Discussion
Our estimates suggest that public library investment yields measurable returns in adolescent reading outcomes, with returns concentrated among lower-income districts and households facing language-access constraints. We do not interpret our results as evidence of a single dominant mechanism; library construction may operate jointly through expanded access to print material, expanded programming, and the provision of supervised after-school space.
Several limitations warrant emphasis. Wisconsin's experience may not generalize to states with different baseline library infrastructure. Our outcome measure captures only one dimension of reading achievement, and longer-run effects on educational attainment and labor market outcomes remain to be examined.
All protocols received approval under Halverson Institute IRB Protocol #2018-0339.
References
- Constantino, R. (2005). Print environments between high and low socioeconomic status communities. Journal of Educational Research, 98(3), 175–183.
- Halverson, E. R., & Pellegrini, M. J. (2019). The Upper Midwest Cohort Study: Design and methodological foundations. Halverson Institute Methods Series, No. 4.
- Krashen, S. (2004). The Power of Reading: Insights from the Research (2nd ed.). Libraries Unlimited.
- Lance, K. C., & Hofschire, L. (2012). School libraries and student achievement: A meta-analytic review. School Library Research, 15, 1–24.
- Lindquist, A. S., & Park, J. (2021). Measuring affective regional identification: Validation of the RAS-12 instrument. Social Indicators Research, 156(2), 519–541.
- Park, J. H., & Halverson, E. R. (2023). After-school programming and adolescent literacy outcomes. Educational Researcher, 52(4), 211–225.
- Wisconsin Department of Public Instruction. (2024). Public Library Annual Report, 2010–2024 [data file].
Suggested Citation
Funding & Disclosures
This research was supported by program-restricted grants administered through the Halverson Institute. The author(s) declare no competing interests. Views expressed are those of the author(s) and do not necessarily reflect those of the institute or its funders.