Is your market about to run out of PTs? Check the number, not the vibe.

What this tool won't tell you, and what it can't fully separate

• Every one of the 51 modeled jurisdictions lands in High or Critical this cycle. Nothing is in Low, and after the August 2026 rescore nothing is in Moderate either. That's a real finding and a real limitation, stated rather than buried: an index whose entire distribution sits in the top two of four tiers is telling you the country is uniformly strained, but it is also not doing much work separating states at the bottom of the range. Read the ordering and the component detail, not the tier label alone. Near-universal 65+ population growth (the last Baby Boomer cohort turned 65 during the 2020–2024 window this index measures) combined with thin DPT pipelines relative to the fixed anchor puts most of the country in the upper tiers. Low and Insufficient Data remain real, reachable tiers built for future-year comparability. They're just not populated today.
• Demand pressure and trend divergence aren't fully independent. The 2020–2024 realized 65+ growth rate drives the demand component directly and, as the subtracted term, most of trend divergence too, closer to 50% combined weight on one underlying signal than the nominal 35%+15%. Trend divergence's real observed range is narrow, contributing at most a few points either way. Read it as a minor adjustment, not an independently strong signal.
• The pay component was rebuilt on 2026-08-04, and the old one was measuring the wrong thing. Version 1 scored PT median wage divided by statewide average pay across all industries. Auditing it found that the ratio was driven almost entirely by its denominator: it correlated −0.93 with statewide average pay and only −0.15 with PT pay itself. In plain terms, it was roughly 86% a measure of how little a state's general workforce earns. Mississippi, West Virginia and Arkansas scored high on "PT wage premium" because everyone else in those states earns little, not because physical therapists were being bid up. It now compares each state's PT median wage to the national PT median ($102,760, BLS May 2024), which is what the component always claimed to measure. It correlates −0.99 with PT pay. Three tiers changed: DC Moderate to High, Montana and Washington High to Critical. The audit, both candidate fixes, and the full cost in tier changes are published in the build record rather than summarized here.
• Pay position is a weak proxy for scarcity in either direction, and it carries only 10%. Low PT pay can mean employers cannot compete for a mobile licensed workforce, or it can mean the local cost of living is low and the market clears fine. This index reads it the first way. That is a judgment call, not a finding, and it is why the component is weighted at a tenth rather than a third.
• DPT graduates are counted where the program is located, not where the graduate goes on to practice. This is a pipeline-production metric, not a guarantee of in-state supply. New PTs are professionally mobile.
• "No in-state DPT program" (Alaska, Wyoming) is a real, confirmed zero, not a data gap. That's different from Insufficient Data, which is reserved for a genuinely missing or suppressed reporting cell. No state hit that case in this pull; the tier stays defined for a future cycle where one might.
• No precise score to two decimal places, no "years to crisis" countdown. That's false precision this annual, state-aggregate public data can't support.