Both Harvard and MEDSL agree on totals, and they are missing about 50% of the votes cast in the county. The validation dataset reports results for 343 precincts, and we have results for 281. In several of the precincts for which we have results, we still have undercounts — the discrepancy does not seem to follow any pattern / be concentrated in certain municipalities.
Both Harvard and MEDSL agree on totals, and they are missing about 50% of the votes cast in the county. The validation dataset reports results for 343 precincts, and we have results for 281. In several of the precincts for which we have results, we still have undercounts — the discrepancy does not seem to follow any pattern / be concentrated in certain municipalities.