I have a total n00b question. Why would this be a “constant”? Wouldn’t different galaxies and different matter in the universe expand at different rates, and be an acceleration/deceleration, where one observation is the derivative or velocity of that one entity being observed?
A natural follow-up to your question might be: "If everything is expanding, then wouldn't the ruler itself be expanding, so the expansion becomes unobservable?"
I'm not a physicist, but from my understanding, the situation is a bit more complicated than the phrasing in your question suggests.
Observation #1: The light from far-away galaxies is redshifted (spectral lines are a bit off from where we'd expect them to be). This suggests that these galaxies are moving away from us. The farther away the galaxy, the more it is redshifted. This suggests that the farther away the galaxy, the faster it is moving. Observations indicate that the recession speed is directly proportional to distance.
This observation is consistent with general relativity, which suggests an expanding universe with homogeneous mass.
But on a smaller scale, gravitational binding somehow takes over, and on even smaller scale, things like electromagnetic and nuclear interactions start having a greater impact, and that's why the Milky Way isn't itself expanding. For that matter, even Andromeda (0.8 Mpc) is too close to be affected by Hubble-style expansion, which only becomes observable at the multi-megaparsec scale.
Hubble found that the recessional velocity of a galaxy is proportional to its distance. The proportionality constant is called the Hubble constant.
It's a bit of a misnomer though, because it's only constant through space, not time. At the time of discovery it was assumed to be constant in time, too.
The irony of course being that Hubble's original data is, at least in modern scientific analysis, very well compatible with "no expansion". There is _a lot_ of community flavour involved in how we talk about measurement results
