Search EdWorkingPapers by author, title, or keywords.
Lottery-based identification strategies offer potential for generating the next generation of evidence on U.S. early education programs. Our collaborative network of five research teams applying this design in early education and methods experts has identified six challenges that need to be carefully considered in this next context: 1) available baseline covariates may not be very rich; 2) limited data on the counterfactual; 3) limited and inconsistent outcome data; 4) weakened internal validity due to attrition; 5) constrained external validity due to who competes for oversubscribed programs; and 6) difficulties answering site-level questions with child-level randomization. We offer potential solutions to these six challenges and concrete recommendations for the design of future lottery-based early education studies.
The Revista del Centro de Estudios Educativos, numero 3, 1971 included an early Carnoy article on the economics of education: “Un enfoque de sistemas para evaluar la educación, ilustrado con datos de Puerto Rico.” The article used a unique data set that had student test scores, students’ family background characteristics, and information about teachers and other school inputs for about one-third of all students in Puerto Rican schools to estimate relations between teacher characteristics and student test scores controlling for students’ social class, gender, and whether the school was urban or rural. Such data sets were rare in the late 1960s, and so were attempts to understand how education systems worked to produce student learning outcomes—that is, to improve the quality of education.
There is a lot to criticize in the empirical analysis in that early article, but it does show that there was considerable concern about the quality of education in Latin America even back in 1971. That concern has grown greatly in the past fifty years as countries in the region have expanded their educational systems to provide an increasing proportion of youth with secondary schooling and higher education. With that expansion, there has been a shift in focus from policies concerned with access to schooling to policies concerned with improving the quality of schooling (UNESCO, 2005).
Two factors have contributed to this shift. The first is research claiming that quality of education, as measured by international test scores, is a better predictor of economic growth than the number of years of schooling in the labor force (Hanushek and Kimko, 2000; Hanushek and Woessman, 2008). The second is the increase in testing itself, both at the national and international levels. Student test results are being used increasingly to pressure national and local educational systems, schools, and individual teachers to have their students do better on the tests (OECD, 2013). League tables comparing schools, local school districts, regions, and nations against others are now a regular feature of educational politics in many countries of the world. To some extent, international test scores are becoming important enough to affect government legitimacy.
We study the effects of increased school spending in rural American school districts by leveraging the introduction and subsequent expansion of Wisconsin’s Sparsity Aid Program. We find that the program, which provides additional state funding to small and isolated school districts, increased spending in eligible districts by 2% annually and that districts primarily allocated funds to areas with low baseline budget shares. This increased spending has little effect on standardized test scores, but modestly increases college enrollment and completion for students with a low likelihood of attending or completing college.
Does relaxing strict school discipline improve student achievement, or lead to classroom disorder? We study a 2012 reform in New York City public middle schools that eliminated suspensions for non-violent, disorderly behavior, replacing them with less disruptive interventions. Using a difference-in-differences framework, we exploit the sharp timing of the reform and natural variation in its impact to measure the effect of reducing suspensions on student achievement. Math scores of students in more-affected schools rose by 0.05 standard deviations relative to other schools over the three years after the policy change. Reading scores rose by 0.03 standard deviations. Only a small portion of these aggregate benefits can be explained by the direct impact of eliminating suspensions on students who would have been suspended under the old policy. Instead, test score gains are associated with improvements in school culture, as measured by the quality of student-teacher relationships and perceptions of safety at school. We find no evidence of trade-offs between students, with students benefiting even if they were unlikely to be suspended themselves.
The effect of school closures in the spring of 2020 on the math, science, and reading skills of secondary school students in Poland is estimated. The COVID-19-induced school closures lasted 26 weeks in Poland, one of Europe's longest periods of shutdown. Comparison of the learning outcomes with pre- and post-COVID-19 samples shows that the learning loss was equal to more than one year of study. Assuming a 45-year working life of the total affected population, the economic loss in future student earnings may amount to 7.2 percent of Poland’s gross domestic product.
We provide evidence that graduated driver licensing (GDL) laws, originally intended to improve public safety, impact human capital accumulation. Many teens use automobiles to access both school and employment. Because school and work decisions are interrelated, the effects of automobile-specific mobility restrictions are ambiguous. Using a novel triple-difference research design, we find that restricting mobility significantly reduces high school dropout rates and teen employment. We develop a multiple discrete choice model that rationalizes unintended consequences and reveals that school and work are weak complements. Thus, improved educational outcomes reflect decreased access to leisure activities rather than reduced labor market access.
Predictive analytics are increasingly pervasive in higher education. However, algorithmic bias has the potential to reinforce racial inequities in postsecondary success. We provide a comprehensive and translational investigation of algorithmic bias in two separate prediction models -- one predicting course completion, the second predicting degree completion. Our results show that algorithmic bias in both models could result in at-risk Black students receiving fewer success resources than White students at comparatively lower-risk of failure. We also find the magnitude of algorithmic bias to vary within the distribution of predicted success. With the degree completion model, the amount of bias is nearly four times higher when we define at-risk using the bottom decile than when we focus on students in the bottom half of predicted scores. Between the two models, the magnitude and pattern of bias and the efficacy of basic bias mitigation strategies differ meaningfully, emphasizing the contextual nature of algorithmic bias and attempts to mitigate it. Our results moreover suggest that algorithmic bias is due in part to currently-available administrative data being less useful at predicting Black student success compared with White student success, particularly for new students; this suggests that additional data collection efforts have the potential to mitigate bias.
As affirmative action loses political feasibility, many universities have implemented race-neutral alternatives like top percent policies and holistic review to increase enrollment among disadvantaged students. I study these policies’ application, admission, and enrollment effects using University of California administrative data. UC’s affirmative action and top percent policies increased underrepresented minority (URM) enrollment by over 20 percent and less than 4 percent, respectively. Holistic review increases implementing campuses’ URM enrollment by about 7 percent. Top percent policies and holistic review have negligible effects on lower-income enrollment, while race-based affirmative action modestly increased enrollment among very low-income students. These findings highlight the enrollment gaps between affirmative action and its most common race-neutral alternatives and reveal that available policies do not substantially affect universities’ socioeconomic composition.
The Cobb Teaching & Learning System (CTLS) is a digital learning initiative developed for and by the Cobb County School District (CCSD) in Georgia. CTLS became a crucial initiative used by the district to maintain student academic progress during the COVID-19 pandemic. Adopting a mixed-methods approach, this case study seeks to analyze CTLS’s design and implementation, focusing on digital transformation and professional collaboration within CCSD. This case study highlights how CCSD maintains complete ownership in a customized digital learning initiative supported by technology providers.
CTLS’s success comes from its strategic partnership with external technology providers, most notably EdIncites, commitment to professional collaboration, investment in novel technologies, and focus on real-time data. Looking at district-by-district comparisons, Cobb’s level of achievement and learning recovery resembles that of higher performing suburban districts in Georgia as opposed to its closest geographically and demographically comparable peers. Furthermore, 2019-2022 testing data indicates that all GA Milestone End-Of-Course proficiency percentages have already exceeded a 2014 baseline. This suggests that CTLS played a central role in CCSD’s successful recovery after the COVID-19 pandemic.
The overall response to the digital learning initiative from the end users that it is intended to serve has also been overwhelmingly positive. The initiative is now well-positioned to broaden learning opportunities across all schools and improve communication with parents and other stakeholders. CCSD’s experience in scaling CTLS offers useful lessons for districts that are ready to launch and to own their transformative digital learning environment.
Faced with decreasing funds and increasing costs, a growing number of school districts across the United States are switching to four-day school weeks (4DSWs). Although previously used only by rural districts, the policy has begun to gain traction in metropolitan districts. We examine homeowner, teacher, and student outcomes in one of the first metropolitan school districts to adopt the 4DSW. We find 2 to 4 percent home price declines relative to surrounding school districts, a 5 percent decrease in teacher retention for experienced teachers, and a 0.2 to 0.3 standard deviation decrease in student test scores. These results suggest the decision to adopt a 4DSW in a metropolitan setting should not be taken lightly.