One thing to keep in mind is, if these are just the "registered voters" is there a way to verify how many of these names actually voted? It's possible that some large proportion of these people passed away a few years ago and their names haven't yet been taken off the list here. So if the number is high for that reason then it could be within the margin of error for my estimate here.

A random sampling of how many did vs. didn't vote could give a useful ratio to show whether this was abused. I'd say if fewer than 15% of these names have a vote attached then it's less likely to be fraud.

(Based on 4300 estimated alive out of 14500 registered = 30%, then subtract some percentage for turnout: I can't estimate what proportion of centenarians are likely to vote, but it's likely to be less than the general population.)