How fast is the universe expanding?

Cosmology has one number more famous than the rest: the Hubble constant, the current expansion rate of the universe. Measure it by starting from the infant universe, the microwave afterglow of the Big Bang, and running the standard cosmological model forward, and you get about 67 kilometers per second per megaparsec. Measure it directly in today’s universe, using stars of known brightness to build a distance ladder to exploding stars in other galaxies, and you get about 73. The two error bars stopped overlapping years ago. By the mid-2010s the discrepancy had a name, the Hubble tension, and it has since sharpened rather than dissolved.

Why it matters

Both methods rest on superbly tested foundations, so the disagreement means one of two interesting things. Either a subtle systematic error lurks in a generation of precision measurements, which would be a masterclass in how careful science can still fool itself, or the standard model of cosmology, the framework behind nearly every statement about the universe’s age, contents, and fate, is missing an ingredient. The tension is currently the most credible crack in a model that has otherwise passed every test for a quarter century.

What has been tried

The error-hunting has been relentless. The local measurement’s weakest links, the calibration of Cepheid variable stars in crowded star fields, were rechecked with the James Webb Space Telescope, which resolved the crowding and confirmed the earlier photometry. Alternative ladders using different stars give values that land in between, muddying rather than settling. On the early-universe side, independent data beyond the microwave background reproduce the low value. Theorists have proposed new physics to reconcile the two, most prominently early dark energy, a transient component boosting the young universe’s expansion, along with new relativistic particles or modified gravity; each proposal fixes the tension only by straining some other measurement. Referees are coming: gravitational-wave standard sirens offer a distance measure with entirely independent systematics, but remain too few to decide.

Where the edge is

The disagreement stands at a statistical significance that would count as a discovery if it were a particle, yet the field remains genuinely split between hidden error and new physics, with no proposal in either category commanding consensus.

What would count as an answer

Either a culprit systematic, identified and shown to move one measurement onto the other, or an independent method, sirens most likely, landing decisively on one value and a model that accommodates it. The universe expands at one rate; we hold two.