County Clerk
Elections Study Group Report
Intent of the VoterLegal Mandate
Central vs Precinct
III. Problem Background
In 1988, County Clerk Dana DeBeauvoir formed an elections study group to recommend a replacement for the over 20-year-old and failing punchcard ballot counting system. After more than a year of work, the 1988 study group recommended the purchase of a new, optical scan central ballot counting system. In 1990, Travis County purchased the current optical scan voting system from American Information Systems (AIS) at a cost of approximately $680,000. The optical scan technology was chosen because of its ease-of-use for voters, ability to produce fast and accurate election returns, and relatively low purchase price. [For more explanation of the decision process in 1990, see the Travis County Elections Task Force Report of Findings and Recommendations, February, 1990.]
With the purchase of the AIS Optical Scan System, a new community standard of producing substantially complete returns by the 10:00 PM news broadcast was set. The AIS system has worked well for Travis County voters over the last decade. However, issues have arisen during the past several years that challenge the efficiency and capacity of the current system. These factors include the "intent of the voter" audit requirement, ADA compliance, voter population growth, joint election trends, closely contested local races, and election night ballot box deliveries.
Intent of the Voter Audit
Elections historically were conducted using paper ballots that were tallied or counted in the polling place by the election judge and clerks. At the end of the day, a precinct return sheet was prepared and delivered to thecounty clerk for accumulation into the countywide report. In the 1950s, lever machines were introduced as the first mechanical voting tool. The advantages included the elimination of human error in the count and the ability of the voter to ensure that his/her vote correctly reflected his/her intent. However, because of their bulk, lever machines were difficult for elections administrators to transfer and store. As technology has progressed, it has slowly changed the way ballots are cast and counted. Now, the vast majority of counties in Texas use some type of electronic ballot counting system.
The main challenge with these electronic systems is that, regardless of how the voter marks his/her ballot, the accuracy of the count depends upon the machine's ability to correctly count the voter's intent. For those machines not designed to interpret the vast number of ways a voter might communicate his/her intent, the Texas Election Code as well as the Secretary of State demand that a human intercede.
Legal Mandate
When using an electronic ballot counting system, sections 65.009 and 121.001 of the Texas Election Code (T.E.C.) as well as Secretary of State (S.O.S.) directive memorandums (see appendix) require election officials to examine every voted ballot to ensure the scanning machine will count a voter's ballot choices accurately. Failure by the voter to mark a ballot according to the instructions does not invalidate that vote. In practice, if a voter marks his/her ballot contrary to written instructions and in such a manner that the optical scan machine cannot read the ballot or interpret a voter's mark as the voter clearly intended, a resolution process must be conducted to ensure the voted ballot is counted accurately. Election officials are legally obligated to perform a manual examination of all voted ballots to determine that the machine will accurately read the voter choices, however they are written. If a voted ballot is found to be unreadable by the machine, that ballot must undergo resolution, a process through which the ballot will be handled to allow the machine to accurately read and reflect the intended vote.
Precinct Audit vs. Central Audit.
With the advent of centralized electronic ballot counting equipment, whether punchcard or optical scan technology, ballots were no longer tallied or counted in the precinct. The ballots were delivered to the central counting station, a single location where the votes were tallied and accumulated, and final returns prepared by the electronic counting system. One of the drawbacks to this centralized system is that the ballots must be prepared for reading by the electronic counting equipment. Preparing the ballots for counting involves two steps: 1) ensuring that the voter has marked the ballot in a way that the computer will accurately interpret the voter's intent, and 2) stacking and aligning the ballots so they can be directly fed into the counting machine with no additional handling.
Prior to 1996, the reviewing of the voted ballots to ensure clarity of the intent of the voter was conducted in the precinct by the election judge and clerks while the polls were open on election day. Procedures and safeguards for opening the ballot box and examining the voted ballots in the precinct are given in the election code [T.E.C. Section 65.002]. Also during this review, the election officials sorted, straightened, and correctly oriented the voted ballots so these ballots could be directly fed through the scanning machine without any additional preparatory work or special handling at the central counting station. Using the precinct election workers to perform this task during the twelve hours of election day voting provided an efficient use of resources and had a fundamental effect on the speed of the centralized counting process on election night.
In 1993 and again in 1994, the precinct audit procedure came under fire by some members of the community concerned about the security of the voted ballots if the ballot boxes were opened in the precinct. In response to this concern, the countyclerk convened the 1994 Election Study Group to review the practices used during election day. The study group's recommendation was to continue the precinct audit procedure because of its proven efficiency and cost-effectiveness. The group believed the precinct audit contributed greatly to the timeliness of election returns and that early return completion promoted confidence in the election process. [For more explanation of the 1994 Election Study Group findings, see the Travis County Election Study Group Final Report, August 1994.]
However, despite this finding and although no incident of inappropriate handling of ballots at the precinct was ever substantiated, the community concerns over the security of the voted ballots during a precinct audit were inescapable. Thus, in an effort to ensure public confidence in the election process, in 1996 the county clerk chose not to allow election judges to process the voted ballots at the precinct. Locked and sealed ballot boxes were distributed at election day supply pickup, and the ballot boxes were not opened until they arrived at the central counting station for processing.
This change in policy created severe constraints on Travis County's ability to quickly process the voted ballots at the central counting station. However, the county clerk believed the concerns being raised about the precinct ballot audit were spreading fears that could ultimately undermine the community's trust in the conduct of its elections. As the 1994 study group predicted, this change in ballot handling procedures requiring all ballots to be reviewed after they were received at the central counting station resulted in dramatic increases in needed personnel and significantly delayed the final return completion time.
- return to Study Group Members
- continue on to Problem Background, part II
