Curently involved in processing historical election records for parts of the United Kingdom. In, at least, one area there is an interesting situation where a church hall is used as a polling station but the clergy who live in the presbytry across the road have to go to an entirely different polling station to cast their vote. Indeed from a quick look at the data other registered electors who live closer to the first church hall but are required to go to the other polling station seem less likely to vote. Makes me wonder if using OSM data in combination with those historicala records there might be some way to prove/disprove the hypothesis that redirection to a further away polling stations stifles voting.
If this is not the appropriate category to post this then mods please move it where it is appropriate.