Papex

Open problem · Cosmology · Open

Hubble tension

Why do early- and late-universe measurements of the expansion rate disagree?

Summary

Local measurements of the Hubble constant disagree with the value inferred from the cosmic microwave background.

Why it is open

The Hubble constant, H0, is the current expansion rate of the universe, given in kilometres per second per megaparsec (km/s/Mpc) [1]. It can be measured directly in the nearby universe or, once a cosmological model is assumed, predicted from observations of the early universe [1]. The direct route is a distance ladder: geometric distances, such as parallaxes and the maser galaxy NGC 4258, calibrate standard candles such as Cepheid variable stars or the tip of the red giant branch (TRGB) in galaxies that hosted type Ia supernovae, and more distant supernovae extend the scale into the expanding universe [2]. The indirect route fits the standard ΛCDM model, with a cosmological constant and cold dark matter, to the cosmic microwave background (CMB): Planck's 2018 analysis predicts 67.4 ± 0.5 km/s/Mpc [3]. The two routes have disagreed since the first Planck results, published in 2014, and the disagreement is known as the Hubble tension [2]. Its size is quoted in standard deviations (σ).

If the tension "is real, and not due to systematic errors in multiple measurements, it implies that ΛCDM is not a correct description of the Universe", the TDCOSMO collaboration writes; there is "no consensus" on which change to the model could solve it, and only independent methods precise enough to confirm or reject the tension can rule out systematic errors [1]. No body sets a formal bar. A 2026 comparison of proposed models required that a successful one "both reduce the tension and provide a sufficiently improved description of the joint data to justify its additional parameters" [4]. George Efstathiou stresses that the ladder measures the standardised peak brightness of type Ia supernovae, and that the tension between that calibration and the CMB "is insensitive to the actual late time expansion history of the Universe" [2].

What is known

The SH0ES ladder (2022). Using Hubble Space Telescope (HST) observations of Cepheids in the hosts of 42 type Ia supernovae, the SH0ES team of Adam Riess and colleagues reported 73.04 ± 1.04 km/s/Mpc and "a 5σ difference" from the Planck prediction, "with no indication that the discrepancy arises from measurement uncertainties or analysis variations considered to date" [5].

JWST cross-checks and a dispute (2024–2026). The James Webb Space Telescope (JWST) resolves the crowded near-infrared backgrounds of Cepheids in supernova hosts [6]. Comparing early JWST samples, Riess and colleagues found HST Cepheid distances consistent with JWST Cepheid, TRGB and J-region asymptotic giant branch (JAGB) distances, with mean differences of about 0.03 magnitudes, "far smaller than the 0.18 mag 'Hubble tension'" [7]. In 2025, combining JWST Cepheids in 19 hosts of 24 supernovae with the HST data, they reported 73.49 ± 0.93 km/s/Mpc, or 73.18 ± 0.88 with 35 TRGB calibrations added, about 6σ above the ΛCDM expectation from the CMB [6]. The Chicago–Carnegie Hubble Program (CCHP) of Wendy Freedman and colleagues finds lower values. Its 2025 status report gives 70.39 ± 1.22 (stat) ± 1.33 (sys) ± 0.70 (supernovae) km/s/Mpc from the TRGB with 24 supernova calibrators, and 68.81 (TRGB) and 67.80 (JAGB) from its JWST data alone, results it calls "consistent with the current standard Lambda cold dark matter (ΛCDM) model, without the need for the inclusion of additional new physics" [8]. Riess and colleagues attribute the gap of about 2.5 km/s/Mpc between the CCHP JWST sample and the full HST sample to "differences in sample selection" [7]. In 2026 Taylor Hoyt, Freedman and colleagues found about 69 km/s/Mpc, with a spread of 0.6, for their 11 JWST-calibrated supernovae across four published calibrations; with the sample expanded to 24, the values based on the Pantheon+ and Carnegie Supernova Project (CSP-II) analyses rose by 2.0 (3.1σ) and 0.8 (1.4σ) km/s/Mpc [9].

A combined local value (2025–2026). In the H0 Distance Network (H0DN), experts voted before any calculation on which first-rank indicators would define a baseline, then combined the measurements with covariance weighting. The baseline is 73.50 ± 0.81 km/s/Mpc, 7.1σ from the early-universe ΛCDM value of 67.24 ± 0.35 and 5.0σ from DESI's baryon acoustic oscillations (BAO) with Big Bang nucleosynthesis; removing either Cepheids or the TRGB has "a minimal effect on the central value", and replacing supernovae with galaxy-based indicators changes H0 by less than 0.1 km/s/Mpc [10]. Efstathiou notes that "there is not yet a consensus on the distance ladder measurements": CCHP reports lower values and systematic differences between methods, and "Each of these points is contested by H0DN" [2].

The early universe without Planck (2025). Calibrated with Big Bang nucleosynthesis, DESI's BAO measurements give 68.51 ± 0.58 km/s/Mpc, a value independent of the CMB anisotropies that does assume standard physics before recombination; in ΛCDM it is in 4.5σ tension with SH0ES [11].

A ladder-free check (2026). Strong gravitational lensing turns the delays between multiple images of a quasar into distances without external calibrators [1]. In an October 2026 preprint, not peer reviewed, the TDCOSMO collaboration reported 71.1 (+2.4/−2.3) km/s/Mpc from 11 lensed quasars, a 3% measurement. It finds that a combination of late-universe probes without supernovae is in 3.7σ tension with Planck, and one without the distance ladder in 2.7σ tension; its own precision "is not quite sufficient to test the Hubble tension independently of all other methods" [1].

Main approaches

Systematic errors in the local ladder. In this view, unrecognised errors in distances or supernova calibration bias the local value high. Strongest evidence: CCHP finds that the SH0ES distance scale and supernova magnitudes have shifted by amounts comparable to the quoted uncertainty, and judges that "a claim of 5σ tension may be premature" [8]; supernova calibrations diverge as the TRGB sample grows [9]. Main difficulty: JWST shows that "photometric bias in lower-resolution HST data does not account for the Hubble tension" [6], the H0DN value barely moves when single indicators are removed or replaced [10], and lensing combinations without supernovae still differ from Planck by 3.7σ [1].

Extra energy before recombination. Early dark energy behaves like a cosmological constant at redshifts above about 3000 and then dilutes; the sound horizon at decoupling becomes smaller, so the H0 inferred from the CMB is larger [12]. Strongest evidence: in the 2026 "H0 World Cup", a preprint comparing 14 models on the same data, early dark energy and early modified gravity did best, moving the inferred H0 towards 70 km/s/Mpc and cutting the residual tension with SH0ES to about 2.5–3.6σ [4]. Main difficulty: early dark energy worsens the tension with large-scale-structure data, and J. Colin Hill and colleagues judged in 2020 that it "appears unlikely to resolve the H0 tension" [13]; no model reduces the tension "to a fully satisfactory level", and none is favoured over ΛCDM without local H0 information [4].

Changes after recombination. Evolving dark energy, or interactions between dark matter and dark energy, alter the recent expansion [4]. Strongest evidence: DESI prefers evolving dark energy over ΛCDM at 3.1σ with the CMB and 2.8–4.2σ with supernovae, so the late expansion history is not settled [11]; see Nature of dark energy. Main difficulty: late-time models "do not improve over ΛCDM" in the World Cup [4], and Efstathiou finds that uncertainty in the late expansion history has no significant impact on the CMB–ladder tension, a conclusion DESI's hint "does not alter" [2].

A local void. The Milky Way would sit in a region about 20% underdense out to 300 Mpc, as near-infrared galaxy counts suggest, and outflows from it would inflate local redshifts [14]. Strongest evidence: in 2025 Indranil Banik and Vasileios Kalaitzidis found that void models fit 42 BAO distance measurements better than a homogeneous Planck cosmology, cutting the overall tension from 3.3σ to 1.1–1.4σ [14]. Main difficulty: galaxy distances from the CosmicFlows-4 catalogue prefer a void of 70 Mpc or less, at most a tenth of the size inferred from galaxy luminosity density, with bulk flows that disagree with CosmicFlows-4 estimates [15]; lensing distances of several gigaparsecs are "immune to any local inhomogeneity" [1].

Open sub-questions

  1. Sample selection. As JWST covers more supernova hosts, do Cepheid, TRGB and JAGB calibrations converge near 73 or near 70 km/s/Mpc [7, 8, 9]?
  2. Supernova analyses. Why do the Pantheon+ and CSP-II calibrations diverge when the TRGB sample grows from 11 to 24 supernovae [9]?
  3. A ladder-free test. Will larger JWST samples of lensed quasars reach the 1.5% precision TDCOSMO projects, and on which side of the tension will they fall [1]?
  4. Early energy and structure. Can an early-energy model cut the residual tension below about 2.5σ while fitting large-scale-structure data [4, 13]?
  5. Calibrate, then test. Which proposals still succeed when the local measurement enters the fit as the supernova peak brightness rather than as H0 [2, 4]?
  6. The void's size. Can one void model fit BAO distances, CosmicFlows-4 velocities and lensing distances together [1, 14, 15]?

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. TDCOSMO Collaboration: Birrer, S., Cappellari, M., Courbin, F., Fassnacht, C. D., Fleury, P., Frieman, J. A. et al. (2026). TDCOSMO 2026: new insights on the Hubble tension from strong lensing time delays. Preprint, arXiv:2610.14881 (version 1, 10 October 2026; submitted to Physical Review Letters), not peer reviewed. Source · Checked 14 Oct 2026
  2. Efstathiou, G. (2026). Late time dynamical dark energy and the CMB-distance ladder tension. Preprint, arXiv:2607.29562 (version 1, 31 July 2026), not peer reviewed. Source · Checked 14 Oct 2026
  3. Planck Collaboration: Aghanim, N. et al. (2020). Planck 2018 results. VI. Cosmological parameters. Astronomy & Astrophysics 641, A6. 10.1051/0004-6361/201833910 · Checked 14 Oct 2026
  4. Schöneberg, N., Poulin, V., Ferrari, A. G., Finelli, F., Lesgourgues, J., Morelli, L. et al. (2026). The H0 World Cup. I. Summary of the baseline group stage results. Preprint, arXiv:2607.13282 (version 1, 14 July 2026), not peer reviewed. Source · Checked 14 Oct 2026
  5. Riess, A. G., Yuan, W., Macri, L. M., Scolnic, D., Brout, D., Casertano, S. et al. (2022). A comprehensive measurement of the local value of the Hubble constant with 1 km s⁻¹ Mpc⁻¹ uncertainty from the Hubble Space Telescope and the SH0ES team. The Astrophysical Journal Letters 934(1), L7. 10.3847/2041-8213/ac5c5b · Checked 14 Oct 2026
  6. Riess, A. G., Li, S., Anand, G. S., Yuan, W., Breuval, L., Casertano, S. et al. (2025). The perfect host: JWST Cepheid observations in a background-free type Ia supernova host confirm no bias in Hubble-constant measurements. The Astrophysical Journal Letters 992(2), L34. 10.3847/2041-8213/ae0ad6 · Checked 14 Oct 2026
  7. Riess, A. G., Scolnic, D., Anand, G. S., Breuval, L., Casertano, S., Macri, L. M. et al. (2024). JWST validates HST distance measurements: selection of supernova subsample explains differences in JWST estimates of local H0. The Astrophysical Journal 977(1), 120. 10.3847/1538-4357/ad8c21 · Checked 14 Oct 2026
  8. Freedman, W. L., Madore, B. F., Hoyt, T. J., Jang, I. S., Lee, A. J. & Owens, K. A. (2025). Status report on the Chicago-Carnegie Hubble Program (CCHP): measurement of the Hubble constant using the Hubble and James Webb Space Telescopes. The Astrophysical Journal 985(2), 203. 10.3847/1538-4357/adce78 · Checked 14 Oct 2026
  9. Hoyt, T. J., Jang, I. S., Freedman, W. L., Madore, B. F., Owens, K. A. & Lee, A. J. (2026). The Chicago Carnegie Hubble Program: improving the calibration of type Ia supernovae with JWST measurements of the tip of the red giant branch. The Astrophysical Journal 1002(1), 11. 10.3847/1538-4357/ae29eb · Checked 14 Oct 2026
  10. H0DN Collaboration: Casertano, S., Anand, G., Anderson, R. I., Beaton, R., Bhardwaj, A., Blakeslee, J. P. et al. (2026). The Local Distance Network: a community consensus report on the measurement of the Hubble constant at ∼1% precision. Astronomy & Astrophysics 708, A166. 10.1051/0004-6361/202557993 · Checked 14 Oct 2026
  11. DESI Collaboration: Abdul-Karim, M. et al. (2025). DESI DR2 results. II. Measurements of baryon acoustic oscillations and cosmological constraints. Physical Review D 112(8), 083515. 10.1103/tr6y-kpc6 · Checked 14 Oct 2026
  12. Poulin, V., Smith, T. L., Karwal, T. & Kamionkowski, M. (2019). Early dark energy can resolve the Hubble tension. Physical Review Letters 122(22), 221301. 10.1103/PhysRevLett.122.221301 · Checked 14 Oct 2026
  13. Hill, J. C., McDonough, E., Toomey, M. W. & Alexander, S. (2020). Early dark energy does not restore cosmological concordance. Physical Review D 102(4), 043507. 10.1103/PhysRevD.102.043507 · Checked 14 Oct 2026
  14. Banik, I. & Kalaitzidis, V. (2025). Testing the local void hypothesis using baryon acoustic oscillation measurements over the last 20 yr. Monthly Notices of the Royal Astronomical Society 540(1), 545–561. 10.1093/mnras/staf781 · Checked 14 Oct 2026
  15. Stiskalek, R., Desmond, H. & Banik, I. (2025). Testing the local supervoid solution to the Hubble tension with direct distance tracers. Monthly Notices of the Royal Astronomical Society 543(2), 1556–1573. 10.1093/mnras/staf1571 · Checked 14 Oct 2026

Page reviewed 14 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in physics. Checked 29 Sep 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Cosmology & astronomy