00 · IN THREE MINUTES
The answer in three steps
- 1The Hubble constant describes today’s expansion rate, but astronomers reach it through evidence chains built at very different cosmic times.
- 2Planck data interpreted with the base model give 67.4 ± 0.5 km/s/Mpc; the SH0ES gives 73.04 ± 1.04 km/s/Mpc.
- 3The mismatch may expose hidden systematic error or missing physics, but it has not yet identified which explanation is correct.
01 · ONE QUANTITY, DIFFERENT ROUTES
One quantity, different routes
For nearby galaxies, recession speed is approximately proportional to distance. The proportionality today is called H₀, the Hubble constant. Its unit, kilometres per second per megaparsec, means that each additional megaparsec of separation corresponds to an added recession rate in the smooth cosmic expansion.
No instrument simply points at “expansion” and reads H₀ from a dial. Every determination combines observations with calibrations and assumptions. The tension is interesting because mature, highly checked evidence chains continue to land in different places even as their stated uncertainties have narrowed.
02 · CLIMBING THE LOCAL DISTANCE LADDER
Climbing the local distance ladder
The local route starts with geometric distances, such as parallax or maser measurements. Those calibrate Cepheid variable stars. Cepheids then calibrate Type Ia supernovae in the same galaxies, and the much brighter supernovae carry the ladder into the smooth Hubble flow.
SH0ES reported H₀ = 73.04 ± 1.04 km/s/Mpc in its comprehensive 2022 analysis. Each rung can introduce crowding, dust, metallicity or calibration effects, so the team cross-checks anchors, instruments and samples. Webb observations have sharpened Cepheid images, testing one prominent concern rather than eliminating every conceivable systematic.
03 · PREDICTING FROM THE INFANT UNIVERSE
Predicting from the infant universe
Planck mapped tiny temperature and polarization patterns in the cosmic microwave background, light released when the universe was about 380,000 years old. Those patterns encode physical scales, matter content and the conditions from which later expansion developed.
To obtain today’s H₀, researchers fit a cosmological model, usually base ΛCDM, to that early-universe pattern and other data. Planck’s 2018 analysis gave 67.4 ± 0.5 km/s/Mpc. It is an extraordinarily precise inference, but it is not a direct local measurement: change the model’s ingredients and the prediction can shift.
The Hubble constant describes today’s expansion rate, but astronomers reach it through evidence chains built at very different cosmic times.
Published measurements remain inconsistent under base ΛCDM and their quoted uncertainties. That fact does not by itself reveal whether the missing piece is systematic error, model assumptions or new physics.
04 · WHY THE BARS DO NOT OVERLAP
Why the bars do not overlap
The difference is much larger than expected from the quoted random uncertainties. In the SH0ES comparison with Planck under base ΛCDM it reached about five standard deviations. That language measures incompatibility under stated models and error estimates; it does not mean a five-sigma discovery of a particular new particle or force.
Independent routes help diagnose the problem. Tip-of-the-red-giant-branch distances, masers, strong-lens time delays, gravitational-wave standard sirens and baryon acoustic oscillations have different strengths and systematics. Their results do not collapse into one perfectly settled number, which is why the field remains active.
05 · ERROR OR NEW PHYSICS
Error or new physics
A mundane resolution could be an unrecognized calibration bias distributed across one evidence chain. A more radical resolution could alter early expansion, recombination, gravity or the contents of the universe. Many proposed models can move H₀, but they must also preserve the excellent fit to the microwave background, galaxy clustering and primordial element abundances.
This makes the tension a stress test rather than permission to add any new ingredient. A successful explanation must improve the full data set without creating a worse disagreement elsewhere. So far there is no consensus model that clears all of those hurdles.
06 · HOW TO READ THE HEADLINE
How to read the headline
“Two expansion rates” is a useful shorthand, but the universe is not literally expanding at two rates in the same place and time. We have two families of inference about one present-day parameter. Their disagreement tells us that at least one part of our observations, calibrations or cosmological description is incomplete.
That is already valuable. Precision science often advances when separate routes should meet and do not. The honest conclusion is neither “new physics confirmed” nor “somebody measured badly,” but a sharply defined inconsistency that new observations can attack.
07 · SOURCES AND EVIDENCE
Sources and evidence
The explanation above follows the evidence chain below. Links open the original source.
- 01Planck Collaboration · Planck 2018 results VIPRIMARY CONSORTIUM RESULT ↗
This primary consortium result anchors a specific claim or limit used in the article.
- 02Riess et al. · A Comprehensive Measurement of the Local Value of H0PRIMARY STUDY ↗
This primary study anchors a specific claim or limit used in the article.
- 03NASA Science · Hubble Constant and TensionOFFICIAL EXPLAINER ↗
This official explainer anchors a specific claim or limit used in the article.
- 04NASA LAMBDA · Hubble ConstantDATA OVERVIEW ↗
This data overview anchors a specific claim or limit used in the article.