Hubble tension

Two good measurements of the universe's expansion rate that stubbornly disagree.

Measure the expansion rate locally — with Cepheid variables and Type Ia supernovae — and you get about 73 kilometres per second per megaparsec. Derive it instead from the cosmic microwave background, using the standard model of cosmology to run the physics forward from the early universe, and you get about 67. The error bars have shrunk for thirty years and the gap has not closed. Either one of the measurements has a systematic problem nobody has found, or the standard model is missing something between then and now. It is currently the most interesting unresolved number in cosmology, which is why it keeps appearing in headlines.