Standardized testing faces serious challenges on multiple fronts: the scores correlate more strongly with family income than with academic ability, high school GPA predicts college success far more reliably, and the pressure of high-stakes exams narrows what students learn in school. These aren’t fringe concerns. As of early 2026, 75% of U.S. colleges are test-optional or test-free, and only 5% of institutions on the Common App still require test scores. The shift away from standardized testing is already underway, and the arguments driving it are grounded in decades of research.
Scores Reflect Wealth More Than Ability
The single most damaging critique of standardized testing is that scores track closely with family income. Research from Opportunity Insights, a Harvard-based team, found that children of the wealthiest 1% of Americans were 13 times more likely than children from low-income families to score 1300 or higher on the SAT or ACT. Among students from the bottom 20% of the income distribution, only about a quarter even take the SAT or ACT. Of those who do, just 2.5% score 1300 or higher. Compare that to students from the top 20%, where roughly one in six test-takers hits that mark.
These gaps aren’t explained by intelligence or effort. Wealthier families can afford private tutoring, test prep courses, and repeat sittings. They tend to live in districts with better-funded schools. When a metric this closely tied to household income is used to sort students into college tiers and scholarship brackets, it functions less as a measure of readiness and more as a mirror of economic advantage.
GPA Predicts College Success Far Better
If standardized tests were the best available predictor of who will thrive in college, that might justify tolerating their flaws. But they aren’t. A large-scale study from the University of Chicago Consortium on School Research examined over 55,000 Chicago Public Schools graduates who enrolled in four-year colleges between 2006 and 2009. The finding: high school GPA was five times stronger than ACT scores at predicting whether a student would graduate from college.
The predictive power of GPA held consistent across different high schools, something that was not true for test scores. At many high schools in the study, researchers found no meaningful connection between students’ ACT scores and whether they eventually finished college. Students with GPAs under 1.5 had roughly a 20% chance of graduating college, while those with GPAs of 3.75 or higher had about an 80% chance.
The reason GPA outperforms test scores makes intuitive sense. A GPA reflects months of sustained effort across multiple subjects, class formats, and types of assignments. It captures skills like time management, persistence, and the ability to meet varied academic demands. A standardized test captures performance on a narrow set of skills during a few hours on a single day, and students can prepare for it in ways that don’t translate to actual college readiness.
High-Stakes Testing Narrows What Students Learn
When schools are judged by test results, and when student futures hinge on scores, classroom instruction bends toward tested subjects. This phenomenon, called curriculum narrowing, has been extensively documented. Roughly 70% of studies on the topic have found that teachers are directed to increase instructional time in reading and math while cutting time from other content areas.
The data on lost instructional minutes is striking. After the No Child Left Behind Act tied school funding to test performance, weekly instructional time in elementary schools dropped sharply in non-tested subjects. Social studies lost 31% of its weekly minutes, falling from 239 to 164. Science dropped 33%, from 226 to 152 minutes. Art and music fell 35%, from 154 to 100 minutes per week. Physical education saw a similar 35% decline.
Teacher surveys reinforce these numbers. About half of teachers reported that less instructional time and fewer resources were being spent on art and music. Forty percent said the same about foreign language instruction. The cuts hit hardest in lower-income, lower-performing schools, where mandated scheduling for reading and math left almost no room for history, civics, or the arts. Students in those schools, already disadvantaged by fewer resources, received a narrower education as a direct consequence of the testing regime meant to help them.
Testing Stress Undermines the Scores Themselves
Standardized tests are supposed to measure what students know. But research suggests they also measure how students respond to pressure, and that stress response can suppress performance in ways that make the scores less valid.
A study led by researchers at Northwestern and Texas A&M measured cortisol, a stress hormone, in students’ saliva during testing and non-testing weeks at a charter school network in New Orleans. On average, students had 15% more cortisol in their systems during the homeroom period before a high-stakes test compared to non-testing days. Students whose cortisol spiked the most tended to perform worse than their classwork and non-high-stakes test performance would predict. Some students showed the opposite pattern, with cortisol dropping on test days, which researchers described as the body disengaging from an overwhelming task. That response was also linked to lower scores.
The stress effect was not distributed evenly. Students from the most disadvantaged neighborhoods, those with the highest rates of poverty and crime, experienced the largest cortisol changes. Their scores were, in the researchers’ assessment, the least valid measures of what those students actually knew. This creates a compounding problem: the students most likely to be harmed by how scores are used are also the students whose scores are least accurate.
Alternatives Already Exist
Abolishing standardized testing doesn’t mean abolishing assessment. Several frameworks already offer ways to evaluate student learning without the drawbacks of bubble-sheet exams.
Performance-based assessment asks students to construct answers, produce products, or carry out activities in response to problems rooted in real experience. These range from journal entries and research reports to lab experiments and creative performances. The Institute of Education Sciences highlights tools like Virginia’s Quality Criteria Review Tool, which provides a structured framework for building rigorous performance assessments that prioritize student engagement and authentic demonstration of knowledge. The Stanford Center for Assessment, Learning and Equity (SCALE) offers a full assessment system for K-12 educators, and the Performance Assessment Resource Bank provides ready-made tasks along with supporting research. A typical performance-based assessment takes two to four weeks to implement, giving students time to demonstrate deeper understanding rather than recall under pressure.
Portfolio-based systems, where students compile their best work across a semester or year, offer another path. These collections can be evaluated against clear rubrics and provide a richer, more complete picture of a student’s abilities than any single test day. Mastery-based models, which track whether students have demonstrated competence in specific skills before moving on, focus on learning rather than ranking.
What Abolition Would Actually Change
Removing standardized testing from college admissions would eliminate a barrier that disproportionately filters out low-income students while adding little predictive value. The rapid adoption of test-optional policies suggests many institutions have already reached this conclusion. For K-12 education, moving away from high-stakes testing would give schools room to restore instructional time in subjects that have been squeezed out for two decades. It would reduce the incentive to teach to the test and allow teachers to use assessment methods that capture a broader range of student abilities.
None of this means accountability disappears. Schools still need ways to measure student progress and identify where support is needed. But the current system, where a single timed exam can shape a student’s trajectory while reflecting their zip code more than their potential, has costs that increasingly outweigh its benefits. The question isn’t whether we need assessment. It’s whether we need this particular kind, and the evidence suggests we don’t.