Analysis of the Connecticut November 2008 post-election audit

The University of Connecticut’s Voting Technology Research Center has recently made available the results of their analysis of the State of Connecticut’s post-election audit of the November 2008 election. The report “Statistical Analysis of the Post Election Audit Data 2008 November Elections” is available here.

Here’s the report’s summary:

The University of Connecticut Voting Technology Research (VoTeR) Center received the data gathered in the post-election audit performed in the State of Connecticut following the November 2008 election. The audits involved the randomly selected 10% of the districts and the audit returns were conveyed by the Office of the Secretary of the State (SOTS) to the VoTeR Center on December 3rd, 4th and 18th of 2008. The original audit data contained 1311 records, where each record represents information about a given candidate: date, district, machine seal number, office, candidate, machine counted total, hand counted total of the votes considered unquestionable by the auditors, hand counted total of the votes considered questionable by the auditors, and the hand counted total, that is, the sum of undisputed and questionable ballots.

The VoTeR Center’s initial review of audit reports prepared by the towns revealed a number of returns with unacceptably high unexplained differences between hand and machine counts and also revealed substantial discrepancies in cases of cross-endorsed candidates (i.e., candidates whose names appear twice on the ballot because they are endorsed by a minor party). As a result, the SOTS Office performed additional information-gathering and investigation and, in some cases, conducted independent hand-counting of ballots. The resulting information was conveyed to the VoTeR Center on February 18, 2009. Further information gathering was conducted by the SOTS Office to identify the cause of the moderately large discrepancies, and more importantly, to identify a cause of substantial discrepancies for cross-party endorsed candidates. The resulting information was conveyed to the VoTeR Center on April 3, 2009.

This report presents the results in three parts: (i) the analysis of the original audit records that did not involve cross-party endorsed candidates, (ii) the analysis of the audit records for cross-party endorsed candidates, and (iii) the analysis of the records that were revised based on the SOTS Office follow ups. The analysis does not include 43 records (3.3%) that were found to be incomplete, unusable, or obviously incorrect. In more detail, part (i) deals with 776 (63%) records that were complete and contained no obvious audit errors. Among these, 776 records (94%) show a discrepancy of 5 votes or lower, with 583 records (71%) showing discrepancy of 0 or 1 vote between the machine counts and audit hand counts. There are 49 records that have the discrepancy of more than 5 votes and the largest discrepancy is 9. Part (ii) deals with 301 records (23%) involving cross-party endorsed candidates. As a result of the second SOTS Office follow up, it was confirmed that large discrepancies reported for the cross-party endorsed candidates were due to the fact that the auditors did not correctly assign hand counted votes to the specific party endorsements. We present the analysis of the original 240 (18%) cross-party endorsed records, and 61 (5%) records that were revised by the SOTS Office. Part (iii) deals with the revised audit returns gathered by the SOTS Office. Part (iii) discusses the records that were the subject of the last investigation by the SOTS Office. This analysis was performed on request of the Office of the Secretary of the State.

The main conclusion in this report is that for all cases where non-trivial discrepancies were originally reported, it was determined that hand counting errors or vote misallocation were the causes. No discrepancies in these cases were reported to be attributable to machine tabulation. For the original data where no follow up investigation was performed, the discrepancies were small, in particular, the average reported discrepancy is lower than the number of the votes that were determined to be questionable. For the cross party endorsement, it is important for the auditors to perform hand counting of the votes that precisely documents for which party endorsement the votes were cast, and to note all cases where more than one bubble was marked for the same candidate. The SOTS follow up confirmed that in almost all examined problematic instances the discrepancies were a result of an incorrect hand counting. Therefore, the auditors should be better trained to follow the correct process of hand count audit.

The analysis does not include 42 records (3.2% of 1311) that were found to be incomplete, unusable, or obviously incorrect. This is an improvement relative to the November 2007 elections, where we reported 18% of the records that were unusable.