About 70% of the variation in California school district test scores can be attributed to demographic differences rather than the quality of teaching or programs, according to a new report from the California Policy Center published in 2026. The study ranked all 880 traditional public school districts in the state after stripping away structural advantages like household income and parent education levels, revealing which schools are genuinely overperforming or underperforming relative to expectations. The authors argue that California's accountability systems, including the California School Dashboard, overprioritize raw performance without adjusting for these demographic realities.
The report combined machine learning and regression models to predict expected test results based on factors including median household income, adult bachelor's degree attainment, English learner rates, and free or reduced-price lunch eligibility. Districts were then sorted into five tiers based on how far their actual results deviated from predictions. Of the 880 ranked districts, 69 were classified as Significant Overperformers, 198 as Moderate Overperformers, 328 as Expected Performers, 230 as Moderate Underperformers, and 55 as Significant Underperformers. Garden Grove Unified, which serves roughly 37,000 students with a diverse, high-need population, ranked as the top overperformer, while Campbell Union ranked last. Compton Unified, despite serving a student body over 95% classified as high-need, landed in the top ten overperformers. On the opposite end, Saint Helena Unified—which spent $39,536 per student compared to neighboring Napa Valley Unified's $17,390—underperformed expectations despite its resource advantages.
The study found that underperforming districts are, on average, more rural, have smaller enrollments, lower adult bachelor's degree attainment, and higher shares of English learners compared to overperformers. When grouped by California's economic regions, Orange County and Los Angeles County showed higher concentrations of overperformers, while rural areas like Kern, North State, and Central Coast had substantially larger shares of underperforming districts. The authors validated their rankings against California's Differentiated Assistance status—the state's mechanism for flagging struggling districts—and found that extreme overperformers had lower rates of receiving state assistance. A separate validation using California Distinguished School awards from 2024 and 2025 showed a statistically significant correlation between district rankings and award counts, with a p-value extremely close to zero.
The report argues that the current Dashboard buries demographically adjusted comparisons beneath other measures and doesn't use them to determine accountability. The authors write that researchers have urged the state toward growth and residual-gain models for years, and this report demonstrates that a credible statewide version can be built entirely from data the state already collects. The bulk of California's most extreme underperformers are located in rural areas, a trend that persists even after controlling for demographic factors, pointing to structural challenges that uniform state programs can't address. Districts like Compton Unified—once notorious for academic struggles, state takeover, and financial bankruptcy—demonstrate that high-need populations aren't incompatible with strong performance, while districts like Saint Helena show that abundant funding and small class sizes don't guarantee success.
The authors recommend that California prioritize demographically adjusted comparisons within the Dashboard and accountability systems, provide additional targeted support for rural districts, and invest in ground-level studies of top-tier and bottom-tier schools to understand what works. The report is a cross-sectional snapshot rather than a time series, and the district profiles are suggestive rather than definitively causal. Future work should extend the framework across multiple years and incorporate student-level growth measures where available. The takeaway: stripping away demographic advantages reveals that district effectiveness varies dramatically across California, and the state already has the data to build a fairer accountability system.