The most precise measurement of cosmic expansion ever made still doesn’t add up

The classic approach to measuring cosmic expansion — known as the distance ladder — works by chaining together overlapping techniques, each calibrated against the one below it. First you measure nearby stars using parallax, then variable stars called Cepheids, then red giant stars at the tip of the red giant branch, and finally Type Ia supernovae bright enough to be seen across billions of light-years. Each rung depends on the one beneath it, which means a single miscalibrated step can propagate errors through the entire structure.

The H0 Distance Network (H0DN) Collaboration, formed at the ISSI Breakthrough Workshop «What’s under the H0od?» in Bern in March 2025, redesigned this architecture from the ground up. Rather than a ladder with a single path from bottom to top, they built a network — a web of interconnected distance indicators that overlap at multiple nodes. If one method is off, the others catch it. About forty specialists were brought together, deliberately including groups that had previously disagreed, to build a consensus framework rather than champion any individual technique.

73.50 — the number that won’t budge

The collaboration’s result, published on April 10, 2026 in Astronomy & Astrophysics, places the local Hubble constant at H₀ = 73.50 ± 0.81 km/s/Mpc. That corresponds to just over 1% precision — the tightest direct measurement of the local expansion rate ever achieved.

The number sits in sharp contrast with predictions from the cosmic microwave background, the faint afterglow of radiation left over from the early universe. That approach, anchored in the standard cosmological model, yields a value near 67.2 km/s/Mpc. The two numbers differ by roughly 9%, and that gap is many times larger than the combined measurement uncertainties. It cannot be dismissed as noise.

Stress-testing the result

The collaboration ran an extensive series of robustness checks, systematically removing individual techniques, anchor datasets, or full classes of observations to see how much the final value shifted. In most cases, the central value barely moved. Only when Cepheid variables were removed entirely did it drop meaningfully, to 72.51 km/s/Mpc — still far above the early-universe prediction.

This is perhaps the most consequential finding in the paper. It means the Hubble tension cannot be blamed on a flaw hiding in any single method. The network’s redundancy explicitly rules out that explanation. As the authors state directly, if the tension is real — and the weight of evidence increasingly suggests it is — it may point toward physics that lies outside the standard cosmological model altogether.

Infrastructure behind the measurement

The study drew on data from two key programs within NSF NOIRLab: the Cerro Tololo Inter-American Observatory in Chile and the Kitt Peak National Observatory in Arizona. John Blakeslee of NSF NOIRLab served as a collaborating author. Beyond the institutional contributions, the team made a deliberate choice to publish their data and methods openly, creating a platform that future observatories — including the next generation of wide-field, high-precision instruments — will be able to build upon directly.

What the tension might be telling us

The Hubble tension has persisted for over a decade, and proposed explanations run the gamut: dynamical dark energy whose properties evolve over time, modifications to general relativity at cosmological scales, new physics operating in the early universe before recombination, or some combination thereof. None has emerged as a clear winner.

What this measurement does is close the door on one of the last comfortable explanations — that the tension was a measurement artifact waiting to be corrected. With a community-built, multi-technique network delivering a result stable to better than 1%, that escape route is gone. The universe is expanding faster than our best theory says it should, and now we know it with unprecedented confidence.

That is not a crisis for cosmology. It is an invitation.

© 2026 SKYCR.ORG | Homer Dávila Gutiérrez, FRAS. All rights reserved. Reproduction in whole or in part without express authorization is prohibited. Original source: H0DN Collaboration, Astronomy & Astrophysics, DOI: 10.1051/0004-6361/202557993.