I’ve already talked about how high paternal age can increase the per-year and per-generation rates. There’s another factor that can cause variation in the mutation rate: differences in the age of male puberty. At least that seems likely, from what we know of the mechanism driving the male mutation rate – the 23 replications per year of the sperm-forming cells.
Setting up the spermatogonial cells requires about 30 replications: after puberty, those stem cells continue dividing indefinitely, and mutations accumulate at the rate of about 2 per year. So, all else equal, late male puberty should mean a lower male mutation rate. In modern industrialized (i.e. chubby) populations, puberty hits at age 13. Among Bushmen, it’s 17.
Mothers contribute about 15 mutations, independent of age, while fathers contribute about 25 +2*(g-20) – measured in Iceland, a modern industrial population. For a modern industrial population with an average paternal age of 28 and an average maternal age of 28, generation length is 28 and the average kids inherits 56 mutations. That’s 56 per generation, 2 per calendar year. You’d expect to see fewer in a similar population with later puberty: 8 fewer in a population with puberty four years later. So you’d see an average of 48 mutations per generation (85% of the previous rate) and 1.7 per calendar year. To me this suggests that measurements of mutation rates in modern populations may not be quite the thing you want to in order estimate the long-term mutation rate in humans. Measurement on on some fairly hard-scrabble population might be better.