โ† Back to Deep Dives

The Hubble Tension: Is the Standard Model Broken?

The standard cosmological model โ€” ฮ›CDM, for Lambda Cold Dark Matter โ€” is one of the most successful scientific theories ever constructed. It describes the universe's composition, structure, and history with extraordinary precision. It correctly predicts the CMB's temperature fluctuation pattern, the large-scale structure of galaxies, the abundances of light elements from Big Bang Nucleosynthesis, and dozens of other observations across a vast range of scales and cosmic times. By any reasonable standard, it is an extraordinary achievement.

And yet, for the past decade, two independent measurements of the universe's current expansion rate have been giving different answers. Not different by a little โ€” different by enough that the probability of the discrepancy being a statistical fluctuation is less than one in a million. Both measurements have been checked, rechecked, and improved by teams of physicists operating independently using different methods, different instruments, and different data. The disagreement has not gone away. It has gotten worse.

This is the Hubble tension. It is either the most significant systematic error in modern cosmology, or it is evidence that the standard cosmological model is missing something fundamental. Either answer would be important. The second answer would be revolutionary.


The Hubble constant โ€” what it is and why it matters

The Hubble constant (Hโ‚€) describes how fast the universe is expanding right now. It has units of km/s/Mpc โ€” kilometers per second per megaparsec. A value of 70 km/s/Mpc means that a galaxy 1 megaparsec (~3.26 million light-years) away is receding from us at 70 km/s. A galaxy 2 megaparsecs away recedes at 140 km/s. The recession speed scales linearly with distance โ€” this is Hubble's law, observed by Edwin Hubble in 1929 and now confirmed to extraordinary precision across billions of light-years.

The Hubble constant is one of the most important numbers in cosmology. It sets the age of the universe (roughly 1/Hโ‚€ gives the Hubble time, an approximation of the universe's age). It determines the distance to any galaxy whose redshift we measure. It controls the predicted abundances of dark matter and dark energy required to match observations. Getting Hโ‚€ right is not just a matter of precision โ€” it is foundational to the entire cosmological model.

67.4
Early Universe (CMB)
Measured by Planck satellite from CMB fluctuations. Based on physics at 380,000 years after the Big Bang, extrapolated forward using ฮ›CDM.
73.0
Late Universe (Distance Ladder)
Measured by the SH0ES team using Cepheid variable stars and Type Ia supernovae in the local universe. Direct measurement today.

The two values above โ€” 67.4 and 73.0 km/s/Mpc โ€” are inconsistent at the 5-sigma level. In particle physics, 5-sigma is the threshold for claiming a discovery. Here, it means the two measurements are so far apart that if one is correct, the other is almost certainly wrong โ€” or there is new physics that makes both correct in different contexts.


The two measurements โ€” how they work

The early universe measurement (CMB)

The Planck satellite's measurement of the CMB gives Hโ‚€ = 67.4 ยฑ 0.5 km/s/Mpc. This is not a direct measurement of the expansion rate today โ€” it is the expansion rate inferred from the CMB by assuming the standard ฮ›CDM model and running the equations forward 13.8 billion years. The CMB fluctuation pattern is sensitive to the universe's composition (the amounts of ordinary matter, dark matter, and dark energy), and those compositions in turn determine how fast the universe expands at any given time. The Planck value is extraordinarily precise โ€” a half-percent uncertainty โ€” but it is model-dependent. It is the value Hโ‚€ must have if ฮ›CDM is correct and the CMB measurements are right.

The local distance ladder

The SH0ES (Supernovae H0 for the Equation of State) collaboration measures Hโ‚€ = 73.0 ยฑ 1.0 km/s/Mpc using a multi-rung distance ladder. The first rung uses parallax โ€” direct geometric distance measurements to nearby Cepheid variable stars using the Hubble Space Telescope and Gaia satellite. The second rung uses these calibrated Cepheids to calibrate the distances to nearby galaxies hosting Type Ia supernovae. The third rung uses those supernovae โ€” standardizable candles โ€” to measure distances to galaxies across billions of light-years. The recession velocities of those distant galaxies, divided by their distances, give Hโ‚€. This measurement is direct โ€” it requires no assumption about the cosmological model.

๐Ÿ“œ A Decades-Long Measurement

Edwin Hubble's original 1929 measurement of the expansion rate gave Hโ‚€ โ‰ˆ 500 km/s/Mpc โ€” roughly 7 times too high, because his distance scale was badly miscalibrated. The true value took decades to nail down: by 1970, estimates ranged from 50 to 100. The Hubble Space Telescope Key Project in 2001 pinned it to 72 ยฑ 8. For years, the tension was within measurement uncertainties. As both the CMB and distance ladder measurements improved in precision through the 2010s, the central values converged on their respective estimates while the uncertainties shrank โ€” and the disagreement became statistically significant. Getting more precise made the problem worse, not better.


Systematic errors or new physics?

Could the distance ladder be wrong?

Systematic errors in the distance ladder are the most natural explanation. Cepheid calibration, dust corrections, metallicity effects, and the selection of Type Ia supernovae all involve assumptions that could introduce biases. The JWST has now independently measured Cepheid distances in key calibrating galaxies and confirmed the SH0ES results to within the uncertainties โ€” making a Cepheid calibration error less likely. Alternative distance indicators โ€” the tip of the red giant branch (TRGB), surface brightness fluctuations, gravitational wave standard sirens โ€” give values scattered between 68 and 73, with most clustering around 70โ€“73. No identified systematic error in the distance ladder has been sufficient to bring it into agreement with the CMB value.

Could the CMB measurement be wrong?

Less likely, but not impossible. The Planck CMB measurement is model-dependent โ€” it assumes ฮ›CDM. If ฮ›CDM has a flaw in its description of the early universe, the inferred Hโ‚€ would be wrong. Possible culprits include extra dark radiation in the early universe (additional relativistic species like sterile neutrinos or dark sector particles), early dark energy (a component that behaved like dark energy only in the early universe), or modifications to the primordial power spectrum. These proposals have been extensively analyzed. None resolves the tension completely without creating new tensions elsewhere. The CMB data itself appears internally consistent within ฮ›CDM โ€” the Hubble tension is a tension between the CMB and the local universe, not within the CMB itself.

"If the Hubble tension is real, it is the most important clue to new physics in cosmology since the discovery of dark energy."

New physics

If neither measurement is wrong, new physics is required. The leading proposals include: Early Dark Energy โ€” a component that behaved like dark energy only in the early universe, speeding up expansion before recombination and increasing the inferred Hโ‚€ from the CMB; modified gravity โ€” deviations from general relativity on cosmic scales that alter the expansion history; dark matter interactions โ€” dark matter that decays or interacts with dark energy, changing the universe's expansion rate at intermediate times; and running vacuum energy โ€” a cosmological constant that evolves over time rather than staying constant. Each proposal introduces new free parameters and must be checked against other observational constraints. None is yet broadly accepted as the solution.

โš  The S8 Tension

The Hubble tension is not the only crack in ฮ›CDM. The "S8 tension" is a related discrepancy between the amplitude of matter clustering predicted by the CMB (combined with ฮ›CDM) and the amplitude measured directly by weak gravitational lensing surveys. The CMB predicts more clustering than is observed โ€” at roughly 2โ€“3 sigma significance. This is less severe than the Hubble tension, but the fact that two independent tensions point in the same direction โ€” toward the early universe overpredicting structure compared to direct late-universe measurements โ€” suggests they may share a common origin. A resolution that explains both tensions simultaneously would be much more compelling than one that fixes only the Hubble tension.

๐Ÿค” What would it mean if the standard model is actually broken?

โ–ผ

It would mean that ฮ›CDM โ€” despite its extraordinary successes โ€” is an approximation, valid over the range of conditions it has been tested in but missing some ingredient or process that becomes important at specific epochs or scales. This is not unprecedented in physics: Newtonian mechanics was extraordinarily successful for 200 years before general relativity revealed its limitations. The Standard Model of particle physics is incomplete (it doesn't include gravity, dark matter, or explain baryogenesis). A modified cosmological model would need to reproduce all of ฮ›CDM's successes while resolving the Hubble and S8 tensions. Finding such a model would be a major achievement and would likely reveal new fundamental physics โ€” possibly new particles, new forces, or new properties of space itself. The Hubble tension may be the first hint of the next revolution in our understanding of the universe.

Key Terms

Hubble Constant (Hโ‚€)
The current expansion rate of the universe in km/s/Mpc. The central quantity in the Hubble tension controversy.
Hubble Tension
The ~5-sigma discrepancy between CMB-based (~67 km/s/Mpc) and distance-ladder-based (~73 km/s/Mpc) measurements of Hโ‚€.
ฮ›CDM
Lambda Cold Dark Matter โ€” the standard cosmological model. Lambda = dark energy (cosmological constant), CDM = cold dark matter.
Distance Ladder
A chain of overlapping distance measurement techniques, each calibrated against the previous, used to measure Hโ‚€ in the local universe.
Cepheid Variables
Stars with pulsation periods proportional to luminosity. Key first rung of the distance ladder, calibrated by geometric parallax.
Type Ia Supernova
Standardizable candle used to measure cosmological distances. Key third rung of the distance ladder.
Early Dark Energy
A proposed component active only in the early universe that would increase the CMB-inferred Hโ‚€ and potentially resolve the tension.
S8 Tension
A related discrepancy between ฮ›CDM-predicted matter clustering amplitude and direct weak lensing measurements. May share a common origin with the Hubble tension.