Many Americans believe that the foremost mission of public education is to provide a pathway to success for every student, even in the face of considerable life obstacles. Yet persistent achievement gaps along dimensions of race, income, family education level, and other factors call this earnest expectation into question. A recent CALDER working paper by thirteen scholars—among them Tim Sass, Cory Koedel, David Figlio, Dan Goldhaber, Eric Hanushek, and Steve Rivkin—examines academic mobility in U.S. public schools to determine the extent to which that vaunted pathway to success actually exists for all public school students. By academic mobility, they mean how much students’ ranks in the distribution of academic performance change during their time in school. An education system with high academic mobility would be one where students’ early-grade ranks are less predictive of their later-grade ranks; one with low academic mobility would see students locked into their early-grade ranks for their entire school career.
The researchers use data from seven states—Georgia, Massachusetts, Michigan, Missouri, Oregon, Texas, and Washington—comprising nearly 3 million students. They follow four cohorts of all students in those states between 2005–06 and 2008–09 who have standardized test scores in math and ELA in the third grade (the initial testing grade in states) and track them through high school, where they typically assess progress in English language arts in tenth or eleventh grade (which tends to be the exam with the highest coverage rate administered in a common grade). Students attend traditional district schools and charter schools, although the latter contribute little to the results since there are comparatively few of them in the total sample.
Their four outcomes of interest are eighth grade test performance, high school test performance, on-time high school graduation, and high school graduation within one year of being on time. Their focus is on upward mobility among initially-low-achieving students. They primarily assess absolute upward mobility (as opposed to relative mobility that compares kids to their peers), which is gauged from the 25th percentile of the distribution of initial performance ranks.
The study follows the framework developed by Raj Chetty and colleagues to study intergenerational economic mobility. That mostly means that their test-based mobility metrics are constructed based on percentile rankings in the test distribution at different points in time during the schooling career.
Overall, students’ ranks in the distribution of academic performance are highly persistent during K–12 education; thus, it follows that absolute upward mobility is low. For example, on average across the seven states, a student who starts at the 25th percentile in the third grade can be expected to perform at roughly the 30th percentile by high school.
Among students who begin with a low performance rank, those from more advantaged backgrounds generally have greater upward academic mobility than their peers from less advantaged backgrounds. But despite finding that academic mobility is low on average in U.S. public schools, the researchers do see statistically significant differences in upward mobility across school districts, which tends to be driven by differences in their initial “baseline” of mobility, when students were first tested
They further find that absolute upward mobility is largest in districts serving students from more socioeconomically advantaged backgrounds. For instance, mobility is higher in districts where local-area incomes, education levels, and residential stability are higher; and where more Asian and White families live. Outside of those attributes, district value-added to student achievement, which controls for student characteristics, is also a consistently strong predictor of high upward mobility. Additionally, because high school graduation rates tend not to be discerning (rates are already so high), students’ early career performance rank is a weaker predictor of their likelihood of graduating.
In summary, academic mobility as a whole appears to be limited, at least as defined here. These results underscore existing research on the persistence, and even widening, of achievement gaps that occur during K–12 schooling. Yes, that feels rather defeatist; however, there is one practical implication: Because baseline mobility is a key driver of differences across districts—and districts with high baseline mobility increase progress throughout the achievement distribution—low-performing students tend to experience their largest gains in districts where students generally excel overall. So district leaders should think twice about targeting additional resources and improvement strategies to specific groups and/or specific grade levels, as is often the case. That’s because building a culture of excellence in schools early on—across the board, aimed at all students—could be the rising tide that lifts all boats.
SOURCE: Wes Austin et al., “Academic Mobility in U.S. Public Schools: Evidence from Nearly 3 Million Students,” CALDER Working Paper (March 2023).
