The Frame-Dragging Bound, Priced: What Satellite Laser Ranging Allows the Torsion Sector

I priced the Einstein–Cartan torsion sector against the strongest satellite gravitomagnetism measurement in existence: the LAGEOS / LAGEOS 2 / LARES satellite-laser-ranging determination of frame-dragging at **0.9910 ± 0.02 (systematic)** of the general-relativistic value (Ciufolini et al. 2019, 7 years of LARES ranging plus 26 years of LAGEOS and LAGEOS 2, GGM05S gravity model). The verdict is printed here at full size, in both directions. In Einstein–Cartan theory proper, torsion is non-propagating and algebraically sourced by intrinsic spin density; where the spin density is zero, torsion is zero — not small, zero. A laser-ranged satellite flies in vacuum, so the Einstein–Cartan contribution to its node precession is **exactly null**, and its measured precession is the metric (Lense–Thirring) precession of general relativity, recovered by my own arithmetic at 30.7 milliarcseconds per year for LAGEOS against the published ~30.7 mas/yr. The measurement therefore does **not** test Einstein–Cartan torsion at all. What it does test — and bind hard — is everything else: any non-Einstein–Cartan extension in which torsion, or any anomalous gravitomagnetic coupling, propagates into the vacuum and adds to the node rate is bounded to **|ΔΩ| 1.0 mas/yr on the combined observable (about 2% of the general-relativistic rate) at the level of the quoted systematic error**, with the measured central deviation -0.009 ± 0.020 of the general-relativistic value. Anyone who cites LAGEOS/LARES as evidence for or against Einstein–Cartan torsion is spending a measurement on a theory that makes no claim in that channel. I show the arithmetic, I label every estimate, and I state exactly what is bounded, at what number, and what is not tested.