An analysis of records from the February 3rd Tulsa City Council primary election shows that there may have been 50 precincts where more votes were cast in the primary election than the number of voters who are recorded as having signed the precinct register. This is the problem that was uncovered in precinct 20, where nearly 50 more votes were cast in the Democrat city primary than the number of Democrat voters who signed the register. This discrepancy was bigger than the margin of victory in the race, rendering the result mathematically uncertain, and forcing a rerun of the vote, which will take place on April 6th.
In five precincts, the discrepancy was in the double-digits:
| PRECINCT |
DISTRICT & PARTY |
VOTES CAST |
VOTERS OF
CORRECT PARTY WHO SIGNED IN |
DISCREPANCY |
| 79 |
5
(R) |
211 |
60 |
151 |
| 144 |
2
(R) |
129 |
49 |
80 |
| 20 |
3
(D) |
255 |
207 |
48 |
| 174 |
2
(R) |
264 |
240 |
24 |
| 119 |
2
(R) |
119 |
108 |
11 |
None of the other Council races were close enough for the discrepancy to affect the outcome.
What could explain these gaps between the number of votes and the number of voters?
For the 45 precincts with a small gap between the number signing in and the number of votes recorded at the polling place, it could be that, in transcribing voter IDs from the register to the election board computer, an ID was accidentally skipped. If someone correctly voted at the primary, then switched parties between the election and the end of February, that would also show up as a discrepancy using my method, which depends on the data as entered into the Oklahoma State Election Board database. To confirm the actual number of invalid votes would require looking at the registration books themselves.
It is unlikely that there could have been enough clerical errors to account for the big gaps in precincts 20, 79, and 144. Witnesses have testified that Republicans were given Democrat city primary ballots in precinct 20, and this is borne out by the fact that 207 voters signed the Democrat register, 53 signed the Republican register, for a total of 260. 255 votes were counted at the polling place.
In precinct 79, 60 voters signed the Republican register, 163 signed the Democrat register, for a total of 223. 211 votes were counted at the polling place. Here again it appears that everyone showing up to vote was given a Republican primary ballot.
The situation in precinct 144 is not quite as clear-cut. 206 voters signed the Democrat register, 49 signed the Republican register, but only 129 people voted -- perhaps only 80 Democrats were incorrectly given Republican ballots. My theory is that they started out giving ballots to everyone, but one of the precinct workers thought better of the idea later in the day, or else a voter called them on it.
What does this mean? See the next entry for my analysis.
