US test scores in Reading in 2026?
Based on persistent multi-year declines in reading across grades and districts, I assess a **62%** chance that U.S. reading scores will show a statistically meaningful decline in 2026; limited improvements for younger children reduce but do not erase the risk of further drops.
Analysis
**Stage 1 — Blind Analysis (ignore market prices)**
The long-run empirical signal is clear: reading scores have trended downward for most cohorts and across a very large share of districts for much of the past decade. Multiple recent analyses (NAEP and district-level reporting) show declines predating COVID and continuing through the most recent releases. Broad-based district-level drops and consistent weakness among older students are the strongest data points supporting a further decline in 2026.
Key mitigating signals are narrower but real: the 2025 NAEP long-term trend reported improvement for 9‑year‑olds, and several early-literacy initiatives and policy pushes toward “scientific” reading instruction have gained traction. Those effects are likely to show up first in younger cohorts and in localized pockets, and they often take multiple years to scale to national NAEP-level samples. Measurement noise and the definition of “significant” (statistical vs. operational) add uncertainty: the NAEP’s sampling variability and any changes in administration, scaling, or state-level policies can push year-to-year results across a significance threshold even without large underlying skill changes.
Weighing the evidence: the baseline informed by the multi-year downward trend and the breadth of district declines gives a strong prior toward decline (I would begin with a ~70% prior that 2026 will show a decline). I then downgrade that prior to account for: (a) the real but limited improvement for younger children in 2025; (b) the possibility that targeted early interventions and policy changes will begin to show localized gains; and (c) statistical uncertainty and sampling noise. That adjusts the independent estimate to **~62%** chance of a meaningful decline in 2026.
Probability intuition underlying the point estimate (summary of influences): - Strong downward trend and broad district-level declines: large upward pressure on P(Decline). - 2025 improvement for 9-year-olds and growing early-literacy policy focus: moderate downward pressure. - Measurement noise / definition of significance: adds variance but not a directional reversal.
**Stage 2 — Market Calibration (compare independent view to market prices)**
Current market allocation among the three outcomes is: Significant decrease 47%, No significant difference 41%, Significant increase 14%. My independent estimate (Significant decrease 62%) is materially higher than the market’s 47% for decline — about a 15-point gap. Several reasons could explain that divergence:
- Market participants may be overweighting the visible 2025 long-term trend improvement for younger students and extrapolating it to national outcomes for 2026, producing too much optimism about stabilization. - Traders may be conflating near-term policy attention (press and reform talk) with immediate measurable gains. Education reforms commonly require several years before producing NAEP-scale measurable effects; the market may be giving those reforms premature credit. - Ambiguity about the event wording and what counts as a "decline" (statistical significance threshold, which test/grade/metric) could push bettors toward the middle category (No significant difference) as a default—raising the market price for no-change and reducing the price for a decrease. - Some informed participants may be risk-managing exposure across correlated education markets, or liquidity-seeking traders may be compressing probabilities toward 50/50. The event has modest volume (≈51.6k contracts), so directional trades could be underrepresented.
Calibration implication: if you share my view that the long-term negative trend and district-level breadth dominate short-term policy and cohort counter-signals, the market underprices the chance of a decline. A directional trade favoring "Significant decrease" at 47% would be positive-expected-value relative to my 62% edge, adjusted for execution costs and your confidence.
(Notes on uncertainty and interpretation: The forecast hinges on how "significant decrease" is operationalized and on NAEP/sample noise. My 62% is a probabilistic judgment about a measurable, statistically supported decline on the national reading metric used by the market — it is not a claim about large-magnitude collapse, but rather that results will tilt negative beyond sampling error.)
Arguments
For
- Long-running negative trend: reading scores have drifted down across grades since the mid-2010s, creating strong prior momentum for further decline.
- District-level breadth: reports show reading fell in ~83% of districts with available data — breadth makes a national reversal less likely in a single year.
- Older-student weakness: persistent declines among older students (who dominate NAEP samples for grade-level national reports) weigh toward continued national declines.
- Structural drivers remain unresolved: chronic absenteeism, weak early instruction, and curriculum issues are systemic and unlikely to be fully corrected by 2026.
Against
- 2025 improvement for 9‑year‑olds is a meaningful counter-signal that could presage stabilization as those cohorts age into tested grades.
- Policy attention and scaling of evidence-based early reading interventions could begin producing measurable gains in pockets that affect national sampling.
- Statistical ambiguity and administration/sample noise could leave year-to-year NAEP changes within the margin of error, producing a 'no significant difference' outcome.
- Some state-level policy changes raise standards rather than skills, which can complicate year-to-year trends both upward and downward depending on reporting and participation.
Key drivers
- Long-term NAEP and district-level trend declines in reading across multiple grades
- 2025 NAEP long-term trend improvement for 9‑year‑olds (young-cohort signal)
- Pace and scale of early-literacy interventions and adoption of scientific reading instruction
- Sampling variability, test administration differences, and the statistical definition of "significant" decline
Risk factors
- Ambiguity in how 'decline' and 'significance' are defined by the market (statistical vs. practical)
- NAEP sampling error or changes in test design/administration that could mask or amplify year-to-year movement
- Recent targeted improvements concentrated in younger cohorts that may not yet translate to national NAEP results
- Policy-induced reporting changes (e.g., tougher standards or new cut scores at the state level) that affect apparent proficiency independent of real-skill changes
Scenarios
Best case
Significant decrease occurs and is clear: NAEP and aggregated district measures show a statistically significant national decline in 2026, reinforcing the narrative of a sustained reading recession and prompting renewed large-scale policy responses and federal/state initiatives focused on remediation.
Most likely
A measurable but modest national decline: results show a statistically significant decline concentrated in older grades and many districts, while younger cohorts show pockets of improvement. The headline is a continued reading problem, albeit with early signals of stabilization among the youngest students — prompting targeted rather than wholesale policy shifts.
Worst case
No significant decline (or even increase) is observed: the 2026 national measures fall within sampling error or improve, driven by the cohort effect from 9‑year‑old improvements, strong localized interventions disseminating rapidly, or favorable test-administration factors. This outcome encourages complacency and delays aggressive system-level reforms.
Outcomes
| Outcome | AI | MKT |
|---|---|---|
| Significant decrease | 62% | 47% |
| No significant difference | 28% | 41% |
| Significant increase | 10% | 14% |
More from this day
- PoliticsKalshi2y
Who will Trump pardon?
AI10%MKT46%Edge-36HypedI assess a low but non-zero chance (~10%) that Barron Trump will receive a presidential pardon before Jan 21, 2029 — legally possible (preemptive pardons exist) but politically unlikely given no current charges or investigations.
- HealthKalshi2y
What will the average number of measles cases be during Trump's term?
AI65%MKT30%Edge+35Hidden GemInterpreting the market question as whether the 2025–2028 average will exceed 1,500 measles cases/year, I assess a 65% chance that the four-year average will be >1,500 cases/year based on the already large 2026 total and persistent drivers lowering population immunity.
- PoliticsKalshi18y
Which G7 leader will leave next?
AI60%MKT92%Edge-32HypedI assess a moderately high chance that the UK Prime Minister (Keir Starmer) will be the first of the current G7 leaders to leave office, but not as extreme as the market; I assign 60% to the UK being first, with the remainder split among plausible early exits from France, the US, Italy and Germany.