What If a State’s Test Scores Are Balanced? New Study Looks Beyond Average Achievement

For decades, the central question in educational decentralization research has been deceptively simple: do states that hand more power to local school districts produce higher average test scores? A new study by Bongyoung Kim of Hosan University and Yong Hak Lee of South Korea’s Ministry of Education, published in SN Social Sciences, argues that this question misses something important. Instead of asking only how tall the peaks are, the researchers asked whether the terrain is even — whether a state’s students perform consistently across subjects and grade levels, or whether some areas of achievement tower over others while others lag behind.

The study draws on data from the National Assessment of Educational Progress, the federally administered examination often called the nation’s report card, covering all 50 U.S. states. Rather than collapsing this information into a single average, the authors constructed a standardized measure of achievement-profile dispersion across four assessments: Grade 4 mathematics, Grade 8 mathematics, Grade 4 reading, and Grade 8 reading. Each subject-grade assessment was standardized separately across states before within-state dispersion was calculated, ensuring that the measure reflects a state’s internal balance rather than its raw performance level. A state with a low dispersion score shows a relatively balanced profile across the four assessments; a state with a high dispersion score shows a more uneven one, with strong performance in some areas coexisting with weakness in others.

This distinction matters because average achievement and profile balance are conceptually different things. A state could post impressive mean scores while showing a lopsided profile — excelling in eighth-grade reading, for instance, while fourth-grade mathematics trails. Another state could have modest averages spread evenly across all four assessments. Traditional decentralization research, which focuses on means, would treat the first state as more successful. But the internal balance of achievement may speak to different dimensions of an education system, such as how evenly resources, curriculum, and instruction reach students across subjects and grade levels.

To test whether governance arrangements shape this balance, the authors examined three dimensions of state-defined governance. The first is fiscal discretion: the degree to which local districts control their own revenues and spending decisions, a dimension rooted in the classic fiscal federalism literature going back to Wallace Oates’s 1972 work on fiscal federalism and Charles Tiebout’s 1956 theory of local expenditures. The second is administrative discretion, capturing how much operational autonomy districts have over personnel, curriculum, and organization. The third is punitive state oversight, measuring the extent to which states impose sanctions, takeovers, or other coercive interventions on struggling districts. The authors also tested whether fiscal autonomy interacts with punitive oversight — whether, for example, local fiscal control might behave differently under aggressive state intervention than under a hands-off regime.

The results are striking for what they do not show. In the fully adjusted cross-sectional regression models, none of the three governance dimensions displayed a statistically reliable association with standardized achievement-profile dispersion. Fiscal discretion yielded a coefficient of 0.029 with a p-value of 0.241; administrative discretion yielded 0.042 with a p-value of 0.186; punitive oversight yielded −0.024 with a p-value of 0.549. The fiscal autonomy by punitive oversight interaction was also statistically imprecise, with a coefficient of 0.023 and a p-value of 0.432. In plain terms, once other factors were accounted for, the study found no clear evidence that any of these governance arrangements is associated with how balanced a state’s achievement profile is.

The authors are careful about what these null findings mean. With only 50 state-level observations, statistical power is limited, and the governance indices themselves are provisional exploratory constructs rather than validated instruments. The analysis is explicitly exploratory rather than confirmatory or causal. A null result in a small cross-sectional sample cannot rule out that governance matters; it can only indicate that no association was detectable with the available data and measures. The study’s primary contribution, the authors write, is conceptual and methodological rather than empirical: it separates average achievement from the internal balance of achievement profiles and clarifies how state-created local discretion and state oversight can be examined as distinct dimensions of educational governance.

The conceptual framing draws on a broader literature in public administration concerning structural coherence and institutional resilience. Structural coherence refers to the alignment of institutions, resources, and practices within a system; institutional resilience refers to a system’s capacity to absorb shocks and continue functioning. The COVID-19 pandemic, which produced documented learning losses across education systems worldwide according to multiple meta-analyses cited in the study, has intensified interest in whether certain governance arrangements make systems more resilient. But Kim and Lee caution that these constructs require direct organizational and longitudinal evidence that their cross-sectional design cannot supply. The achievement-profile dispersion measure, they emphasize, is not a direct measure of educational quality, institutional coherence, or resilience — it captures only the extent to which a state’s observed profile is relatively balanced or uneven.

Methodologically, the study is notable for its transparency about limitations. The authors note that the analysis relies exclusively on publicly available, aggregated state-level secondary data, involving no human participants or identifiable information, so institutional review approval was not required. They employed heteroskedasticity-consistent covariance estimators with improved finite-sample properties, following the MacKinnon and White approach, an appropriate choice given the small sample. The data availability statement notes that no datasets were generated or analysed during the current study beyond the publicly available NAEP data, and the research received no external funding.

For policymakers and researchers, the study suggests that the decentralization debate may need new outcome measures. If the goal of educational governance is not only to raise average scores but to produce consistent performance across subjects and grade levels, then the standard toolkit of fiscal autonomy, administrative discretion, and punitive oversight may not be the right lever — or at least not one whose effects are detectable at the state level with current data. Future research, the authors imply, should pursue longitudinal designs, direct organizational evidence, and more refined governance indices before drawing firm conclusions about how state-created discretion and oversight shape the internal balance of student achievement.

The study, published on 20 August 2026 in SN Social Sciences as volume 6, article number 359, was received on 6 May 2026 and accepted on 18 August 2026. It is available open access at the canonical DOI 10.1007/s43545-026-01659-3. The authors declare no competing interests. Kim and Lee wrote the main manuscript text and prepared all figures, and both authors reviewed the manuscript. The corresponding author is Yong Hak Lee of the Ministry of Education in Sejong City, Republic of Korea.

Subject of Research: Educational governance and achievement-profile balance across U.S. states

Article Title: Beyond mean achievement: an exploratory analysis of state-defined governance and achievement-profile balance across U.S. states

Article References: Kim, B., & Lee, Y. H. (2026). Beyond mean achievement: an exploratory analysis of state-defined governance and achievement-profile balance across U.S. states. SN Social Sciences, 6(9), Article 359. https://doi.org/10.1007/s43545-026-01659-3

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01659-3

Keywords: educational decentralization, NAEP, fiscal autonomy, state oversight, achievement dispersion, exploratory analysis, Beyond, mean, achievement, exploratory, analysis, state-defined