Here's a pretty nice graph that shows the length of the day over the last 60-ish years. It looks like it doesn't have the last few years of data, which I found here. One thing to note is just how much daily variability there is, some of which is seasonal or driven by other cycles, but some of which is more or less random. Here is a breakdown of some of the contributions.

It appears that the year to year changes we see are on the order of a few hundred microseconds, so if earthquakes would be expected to have an effect on the order of a microsecond or so, that could certainly be a (smallish) contributing factor, but not enough to account for the entire change.