How the Hubble Tension Exposes Gaps in Our Cosmological Model

The Universe's Expansion Rate Measured Twice — and the Numbers Don't Match

The Hubble constant H₀ describes how fast the universe expands: for every megaparsec of distance between two points in space, the recession speed increases by H₀ kilometres per second. Two independent measurement approaches now give values that disagree at the 5-sigma level — a discrepancy of roughly 7 km/s/Mpc. In physics, 5 sigma is the threshold for claiming a discovery. This disagreement is known as the Hubble tension, and it has persisted and deepened for over a decade despite extensive efforts to find systematic measurement errors.

The tension may be the most important problem in observational cosmology today. Either both measurements contain unidentified systematic errors of extraordinary coincidence, or the standard cosmological model — ΛCDM — is missing physics that alters the expansion history of the universe.

Two Ways to Measure H₀

The measurements come from fundamentally different eras of cosmic history, each with its own methodology and calibration chain.

The distance ladder (late universe). Astronomers measure distances using a hierarchy of calibrators: geometric parallax calibrates Cepheid variable star distances, Cepheids calibrate Type Ia supernova luminosities, and Type Ia supernovae calibrate the Hubble flow at cosmological distances. The SH0ES (Supernovae and H₀ for the Equation of State) team, led by Adam Riess, used the Hubble Space Telescope and James Webb Space Telescope to calibrate this chain and report H₀ = 73.0 ± 1.0 km/s/Mpc as of 2022.

The CMB (early universe). The Planck satellite mapped the cosmic microwave background with exquisite precision. The angular scale of acoustic oscillations in the CMB depends on the expansion history, and fitting the standard ΛCDM model to the CMB data gives H₀ = 67.4 ± 0.5 km/s/Mpc. This value is extrapolated forward from 380,000 years after the Big Bang to the present, assuming ΛCDM is correct throughout.

Could Systematic Errors Explain It?

Systematic errors have been meticulously examined. The JWST results from 2023 confirmed Hubble Space Telescope Cepheid measurements with unprecedented precision, ruling out many potential issues in the distance ladder. Crowded stellar fields, dust extinction, metallicity effects on Cepheid brightness — each has been examined and found insufficient to bridge the gap.

• Tip of the Red Giant Branch (TRGB) stars offer an independent distance calibration bypassing Cepheids entirely. TRGB measurements yield an intermediate value of about 69–71 km/s/Mpc, lying between the two extremes but not definitively resolving the tension.
• Water masers in NGC 4258 provide a geometric distance anchor independent of parallax. This anchor, used to calibrate both Cepheids and TRGB, does not reduce the tension.
• Strong gravitational lensing time delays (H0LiCOW/TDCOSMO) consistently return values near 73 km/s/Mpc, supporting the high-H₀ camp.
• A 2023 reanalysis of the Cepheid calibration by Breuval et al. found H₀ = 72.8 ± 0.9 km/s/Mpc, maintaining the tension.

What New Physics Might Resolve It

If systematics cannot explain the discrepancy, the ΛCDM model is wrong or incomplete. Proposed extensions fall into two categories: modifications to early-universe physics (reducing the acoustic horizon before the CMB epoch) and modifications to late-universe physics (altering expansion since recombination).

Early dark energy is among the most studied proposals. A scalar field that contributes a fraction of the energy density of the universe just before recombination would shrink the acoustic horizon, causing the CMB-inferred H₀ to increase toward the distance-ladder value. However, early dark energy simultaneously worsens another cosmological tension: the S₈ tension, a disagreement between CMB-predicted and weak-lensing-measured matter clustering amplitude.

The S₈ Tension: A Related Crack

The Hubble tension is not the only anomaly in ΛCDM. The S₈ parameter measures the amplitude of matter clustering: higher S₈ means more clustering. CMB-based predictions exceed weak gravitational lensing measurements (from surveys like KiDS and DES) by about 2–3 sigma. This second tension points in the same direction: the real universe may contain less matter clustering than ΛCDM predicts, or dark energy may not be as simple as a cosmological constant.

• KiDS-1000 data give S₈ = 0.766 ± 0.020, versus Planck's prediction of 0.832 ± 0.013.
• The two tensions — Hubble and S₈ — may be connected, or they may arise from independent physics.
• No single model satisfactorily resolves both simultaneously, which constrains the space of viable extensions to ΛCDM.

What Comes Next

The Rubin Observatory Legacy Survey of Space and Time (LSST), the Euclid mission, and the Nancy Grace Roman Space Telescope will measure baryon acoustic oscillations, weak lensing, and type Ia supernovae with vastly improved statistics over the next decade. If the Hubble tension is real and persists at 5 sigma, these surveys will either confirm new physics or identify the systematic that has eluded two decades of scrutiny. The Hubble tension may prove to be the crack through which a new cosmology is revealed.