This is the accessible text file for GAO report number GAO-14-850 entitled 'Elections: Observations on Wait Times for Voters on Election Day 2012' which was released on September 30, 2014. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. United States Government Accountability Office: GAO: Report to Congressional Requesters: September 2014: Elections: Observations on Wait Times for Voters on Election Day 2012: GAO-14-850: GAO Highlights: Highlights of GAO-14-850, a report to congressional requesters. Why GAO Did This Study: Millions turn out to vote in U.S. general elections, and there were reports of long wait times at some polling places on Election Day in 2012. The authority to regulate elections is shared by federal, state, and local officials; however, responsibility for conducting federal elections primarily resides with about 10,500 local election jurisdictions. GAO was asked to examine voter wait times for the November 2012 election. This report addresses (1) the extent to which local election jurisdictions collected data to measure voter wait times and had long wait times on Election Day 2012, and (2) the factors that affected wait times and their impacts across jurisdictions. GAO surveyed officials from a nationwide generalizable sample of 423 local election jurisdictions, excluding jurisdictions with populations of 10,000 or fewer and in the vote-by-mail states of Oregon and Washington, to obtain information on voter wait times (80 percent responded). Estimates from the survey are subject to sampling error and are reported with 95 percent confidence intervals. GAO also interviewed election officials from 47 of 50 states and the District of Columbia to obtain their views on wait time issues. GAO also selected 5 local jurisdictions based on, among other things, demographic characteristics and estimated wait times to examine in more detail their Election Day 2012 experiences. The results from these 5 jurisdictions are not generalizable, but provide insights into jurisdictions' experiences. GAO also reviewed literature on wait times and interviewed 14 election researchers selected based on their work on election wait times. What GAO Found: the basis of GAO's nationwide generalizable survey of local election jurisdictions, GAO estimates that 78 percent (from 74 to 83 percent) of jurisdictions did not collect data that would allow them to calculate wait times, primarily because wait times have not been an issue, and most jurisdictions did not have long wait times on Election Day 2012. Specifically, GAO estimates that 78 percent (from 73 to 83 percent) of local jurisdictions nationwide had no polling places with wait times officials considered to be too long and 22 percent (from 17 to 27 percent) had wait times that officials considered too long at a few or more polling places on Election Day 2012. Jurisdiction officials had varying views on the length of time that would be considered too long—for example, some officials considered 10 minutes too long, while others considered 30 minutes too long. Because there is no comprehensive set of data on wait times across jurisdictions nationwide, GAO relied on election officials in the jurisdictions it surveyed to estimate wait times based on their perspectives and any data or information they collected on voter wait times. Multiple factors affected voter wait times on Election Day 2012, and their impacts varied across jurisdictions. Specifically, GAO's survey of local election jurisdictions, review of wait time literature, and interviews with election officials and researchers identified nine common factors that affected wait times. Figure: Voting Stages and Nine Key Factors That Affected Wait Times on Election Day 2012: [Refer to PDF for image: illustration] Stage: Arrival; Factors at each stage: * Opportunities for voting before Election Day. Stage: Check-in; Factors at each stage: * Type of poll books[A]; * Determining voter eligibility. Stage: Mark and submit; Factors at each stage: * Ballot characteristics; * Amount and type of voting equipment. Crosscutting factors: * Number and layout of polling places; * Number and training of poll workers; * Voter education; * Resource availability and allocation. Source: GAO 2014 survey of local election jurisdictions and analysis of studies on election administration and interviews with election officials and researchers. GAO-14-850. [A] A poll book is a list of registered voters and is used by poll workers to verify voters' registration. [End of figure] The specific impact of these nine factors depended on the unique circumstances in each of the 5 local jurisdictions GAO selected for interviews, leading to targeted approaches for reducing wait times where needed. For example, according to election officials in 2 jurisdictions, lengthy ballots were the primary cause of long wait times. In 1 of these jurisdictions, state constitutional amendments accounted for five of its eight ballot pages on average, and since the 2012 election, a state law was enacted that established additional word limits to such amendments, which officials said could help reduce wait times. Another jurisdiction that had ballots of similar length did not report long wait times. In comments on draft report excerpts, 1 jurisdiction stated that our description of its experiences was accurate. The Election Assistance Commission and 4 other jurisdictions did not have comments. View [hyperlink, http://www.gao.gov/products/GAO-14-850]. For more information, contact Rebecca Gambler at (202) 512-8777 or gamblerr@gao.gov. [End of section] Contents: Letter: Background: On the Basis of Our Survey, We Estimate That Most Jurisdictions Did Not Collect Data for Calculating Wait Times or Have Long Voter Wait Times on Election Day 2012: A Number of Factors Affected Voter Wait Times on Election Day 2012, and Their Impacts Varied across Jurisdictions: Agency and Third-Party Comments: Appendix I: Objectives, Scope, and Methodology: Appendix II: Aggregated Results from Nationwide Survey of Local Election Jurisdictions: Appendix III: Estimates of Average Wait Times by State on Election Day 2012 Based on Nationwide Public Opinion Survey: Appendix IV: Comments from the Supervisor of Elections for Lee County, Florida: Appendix V: GAO Contact and Acknowledgments: Tables: Table 1: Selected Wait Time Data for Election Day 2012 Collected by Jurisdictions Nationwide: Table 2: Reported Wait Times and Characteristics on Election Day 2012 across Five Selected Jurisdictions: Table 3: Population in Each State Group: Table 4: Local Jurisdiction Election Survey Sample Allocation: Figures: Figure 1: Voting Process in Polling Places on Election Day: Figure 2: Direct Recording Electronic Machine, Central Count Optical Scanner, and Precinct Count Optical Scanner: Figure 3: Estimated Percentage of Local Jurisdictions Nationwide That Had Polling Places with Wait Times Officials Considered Too Long on Election Day 2012: Figure 4: Estimated Percentage of Local Jurisdictions' Average Wait Times at Polling Places at Different Times on Election Day 2012: Figure 5: Estimated Percentage of Local Jurisdictions Nationwide That Had Polling Places with Wait Times of Greater than 60 Minutes at Any Time on Election Day 2012: Figure 6: Voting Stages and Nine Key Factors That Affected Voter Wait Times on Election Day 2012: Figure 7: Estimates of Average Wait Times by State on Election Day 2012 Based on Nationwide Public Opinion Survey: Abbreviations: CCES: Cooperative Congressional Election Study: DRE: direct recording electronic: EAC: Election Assistance Commission: HAVA: Help America Vote Act: MCD: minor civil division: SPAE: Survey of the Performance of American Elections: [End of section] United States Government Accountability Office: GAO: 441 G St. N.W. Washington, DC 20548: September 30, 2014: The Honorable Elijah E. Cummings: Ranking Member: Committee on Oversight and Government Reform: House of Representatives: The Honorable Gerald E. Connolly: Ranking Member: Subcommittee on Government Operations: Committee on Oversight and Government Reform: House of Representatives: The Honorable Frederica S. Wilson: House of Representatives: Millions of individuals turn out to vote in U.S. federal elections, and there were reports of voters who waited in long lines to vote at some polling places on Election Day in November 2012. An estimated 153 million individuals were registered to vote in this election, and an estimated 67 percent of Americans who voted did so in the traditional way by casting ballots at their local polling places on Election Day. [Footnote 1] Reports of long wait times at the polls also occurred in prior elections, including the 2008 and 2004 general elections. Long voter wait times have raised concerns because, according to election administration researchers, they may discourage some people from voting or impose hardships on some voters, such as those who cannot afford to miss work or those with disabilities who are physically unable to wait for long periods of time.[Footnote 2] Researchers and others have defined long wait times in different ways, such as 30 minutes or more or 120 minutes or more. For the purposes of our review, we did not identify a specific amount of time as constituting or defining a long wait time; rather, as discussed in our report, we obtained views from state and local election officials, researchers, and others on what they consider to be long wait times. The authority to regulate elections in the United States is shared by federal, state, and local officials. Deriving its authority from various constitutional sources, depending upon the type of election, Congress has passed legislation addressing major functional areas in the voting process such as voter registration and prohibitions against discriminatory voting practices.[Footnote 3] Nevertheless, the responsibility for the administration of state and federal elections resides at the state level, and state statutes regulate various aspects of elections, including registration and Election Day procedures. Within each state, responsibility for managing, planning, and conducting elections is largely a local process and resides with about 10,500 local election jurisdictions nationwide. Since 2001, we have issued a number of reports on various aspects of the election process, such as voting technology used in federal elections and state laws addressing voter registration. In our reviews of both the 2004 and 2000 general elections, we noted that long lines were identified as an Election Day issue by jurisdictions we surveyed. [Footnote 4] In our reports on those elections, we noted some of the jurisdictions that reported long lines were those jurisdictions that, for example, experienced higher than expected voter turnout or had presidential races that were considered close. You asked us to review voter wait times during the November 2012 general election and the factors that contributed to wait times, among other things. This report addresses the following questions: 1. To what extent did local election jurisdictions collect data to measure voter wait times and have long voter wait times on Election Day 2012? 2. What factors affected voter wait times on Election Day 2012, and what were the impacts of these factors across jurisdictions? To address these objectives, we conducted a web-based survey of election officials from a nationally representative stratified random sample of 423 local election jurisdictions, excluding jurisdictions with populations of 10,000 or fewer and jurisdictions in Oregon and Washington.[Footnote 5] Officials from 80 percent of these jurisdictions responded to our survey.[Footnote 6] In stratifying our nationwide sample, we grouped jurisdictions by their 2010 U.S. Census population--small (10,001 to 100,000), medium (100,001 to 500,000), and large (more than 500,000). We surveyed officials about any data their jurisdictions collected related to wait times on Election Day 2012, voter wait times in their jurisdictions on this day, and their views on factors that affected long voter wait times.[Footnote 7] Responses about wait times may be based on officials' perspectives, data, or other information on wait times. We also analyzed responses from the 2012 Cooperative Congressional Election Study (CCES), a survey of U.S. citizens aged 18 and over, to obtain state-level estimates of wait times reported by voters for in-person voting on Election Day 2012.[Footnote 8] To assess the reliability of these data, we reviewed documentation related to the 2012 CCES and interviewed researchers knowledgeable about the survey. We determined that the CCES data used in this report were sufficiently reliable for our purposes. Further, we interviewed state election officials from 47 states and the District of Columbia to obtain such information as the availability of data on voter wait times in their states for Election Day 2012 and their views on policies and procedures that may have affected voter wait times.[Footnote 9] We corroborated the information we gathered through these interviews by reviewing any documentation that these states provided, such as guidance on planning elections and voter wait time reports. We also interviewed local election officials, on-site or by phone, from 5 local jurisdictions--Detroit, Michigan; Hartford, Connecticut; Lee County, Florida; Los Angeles County, California; and Prince William County, Virginia--to perform a more detailed examination of their experiences on Election Day 2012, including how, if at all, they measured wait times; their views on the factors that affected wait times in their respective jurisdictions on Election Day 2012; and their opinions on the specific impacts of these factors, among other things.[Footnote 10] We selected these jurisdictions (1) to reflect variation in geographic location and demographic characteristics and (2) based on our survey results, CCES results, and our review of wait time literature, to include a range of voter wait times and election administration policies and practices. For example, in our survey, 4 of the 5 selected jurisdictions reported having varying extents of long wait times and 1 reported not having long wait times. In each jurisdiction, we interviewed the chief election official; other officials from the elections office; and, if available, individuals who had served as poll workers at polling locations in the jurisdiction on Election Day 2012. While these 5 jurisdictions are not representative of all local election jurisdictions nationwide, officials in these locations provided a range of perspectives on voter wait times and information on how factors affected wait times in practice and allowed us to compare Election Day 2012 experiences across jurisdictions. We corroborated the information we gathered through these interviews by reviewing postelection reports, relevant state statutes, and documentation that these jurisdictions provided to us, such as data relating to voter wait times and poll worker training materials. In addition, we interviewed officials from the Election Assistance Commission (EAC) and 14 researchers and representatives from research organizations in the field of election administration to discuss their perspectives on wait time measurement and voter wait times.[Footnote 11] We selected these researchers and representatives based on our review of voter wait time literature and their expertise and work in this area. The information that we obtained cannot be generalized to other researchers; however, these interviews provided a range of views on such areas as practices for measuring wait times, the frequency of long voter wait times, and factors affecting wait times. We also reviewed relevant literature on voter wait times, such as studies on wait times by various researchers and reports completed or sponsored by state or local governments in our 5 selected jurisdictions.[Footnote 12] A GAO social scientist and a statistician reviewed the studies whose findings we cite in this report and determined that the design, implementation, and analyses of the studies were sufficiently sound to support the studies' results and conclusions based on generally accepted social science principles. See appendix I for additional information on our scope and methodology. We conducted this performance audit from July 2013 to September 2014 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Background: Overview of Election Administration: Election authority is shared by federal, state, and local officials in the United States, but election administration is highly decentralized and varies among state and local jurisdictions. Federal election laws have been enacted that include provisions pertaining to voter registration, protecting the voting rights of certain minority groups, and other areas of the elections process. States regulate various election activities, including some requirements related to these laws, but generally delegate election administration responsibilities to local jurisdictions.[Footnote 13] Federal Roles and Responsibilities: Congressional authority to regulate elections derives from various constitutional sources, depending upon the type of election, and Congress has passed legislation in major functional areas of the voting process, such as voter registration, as well as prohibitions against discriminatory voting practices. For example, the Help America Vote Act (HAVA)--enacted in October 2002--includes a number of provisions related to voter registration, voting equipment, and other election administration activities, and authorized the appropriation of funds to be used toward implementing the law's requirements. [Footnote 14] The act authorized funding for states and jurisdictions to, among other things, meet the act's requirements, including replacing punch card and mechanical lever voting equipment and creating and maintaining a centralized state voter registration database. HAVA also established the EAC, an independent federal agency, to help improve state and local administration of federal elections.[Footnote 15] The EAC is charged with providing voluntary guidance to states regarding implementing certain HAVA provisions and serving as a national clearinghouse and resource for information with respect to the administration of federal elections, among other things. For example, the EAC issued guidelines that identified data that would be helpful in conducting postelection analysis, such as the average wait time for polling place voters by precinct, and best practices for designing ballots and voter information materials. In addition to HAVA, federal laws have been enacted in other areas of the voting process. For example, the National Voter Registration Act of 1993 expanded the opportunities for eligible citizens to apply to register to vote in federal elections by requiring states to allow registration by mail using the federal voter registration form and at state motor vehicle agencies and other specified public agencies. [Footnote 16] Also, the Voting Rights Act of 1965, as amended, contained, among other things, provisions designed to protect the voting rights of U.S. citizens of certain ethnic groups whose command of the English language may be limited.[Footnote 17] Language minority provisions in the act require covered states and covered jurisdictions to provide written materials--such as sample ballots or registration forms--in the language of certain "language minority groups" in addition to English, as well as other assistance, such as bilingual poll workers.[Footnote 18] State and Local Roles and Responsibilities: The responsibility for the administration of elections resides at the state and local levels. States regulate various aspects of elections including, for example, registration procedures, absentee and early voting requirements, and Election Day procedures. Further, states are required by HAVA to implement a single, uniform, centralized, computerized statewide voter registration list to serve as the official voter registration list for conducting all elections for federal office in each such state.[Footnote 19] Within each state, responsibility for managing, planning, and conducting elections is largely a local government process, residing with about 10,500 local election jurisdictions nationwide.[Footnote 20] Some states have mandated statewide election administration guidelines and procedures that foster uniformity among the ways local jurisdictions conduct elections. Others have guidelines that generally permit local election jurisdictions considerable autonomy and discretion in the way they run elections. Although some states bear some election costs, local jurisdictions generally pay for elections. Local jurisdictions have discretion over such activities as training election officials and the purchase of voting technology (if not mandated by the state). Among other things, local election officials register eligible voters and maintain voter registration lists; design ballots; educate voters on how to use voting technology and provide information on the candidates and ballot measures; arrange for polling places; recruit, train, organize, and mobilize poll workers; prepare and test voting equipment for use; count ballots; and certify the final vote count. The Voting Process: Voting before Election Day: States have established alternatives for voters to cast a ballot other than at the polls on Election Day, including absentee voting and early voting. All states and the District of Columbia have provisions allowing voters to cast their ballots before Election Day by voting absentee, with variations on who may vote absentee, whether the voter needs to provide an excuse for requesting an absentee ballot, and the time frames for applying for and submitting absentee ballots.[Footnote 21] Some states also permit registered voters to apply for an absentee ballot on a permanent basis so that those voters automatically receive an absentee ballot in the mail prior to every election without providing an excuse or reason for voting absentee.[Footnote 22] In addition to absentee voting, some states allow early voting. In general, early voting allows voters from any precinct in the jurisdiction to cast their vote in person without providing an excuse before Election Day either at one specific location or at one of several locations. In-person Voting on Election Day: For the purposes of in-person voting on Election Day, election authorities subdivide local election jurisdictions into precincts. Voters generally cast their ballots at the polling places for the precincts to which they are assigned by election authorities.[Footnote 23] Within the polling place, there are three stages in the voting process--arrival, check-in, and marking and submitting ballots--and poll workers have roles and responsibilities associated with each of them.[Footnote 24] Figure 1 describes the three stages in the voting process. Figure 1: Voting Process in Polling Places on Election Day: [Refer to PDF for image: illustration] Arrival: Voters arrive at polling place. Check-in: Poll workers verify voters' registration using voter lists and in some cases requesting voters provide required identification. Mark and submit: Electronic ballots: Ballot is submitted electronically in voting booth; Paper ballots: Voters submit ballots to be counted at the polling place or to be sent to a central tabulation location for counting. Source: GAO 2014 survey of local election jurisdictions and analysis of studies on election administration and interviews with election officials and researchers. GAO-14-850. [End of figure] * Arrival. Poll workers manage the arrival of voters, which may include tasks such as greeting and directing voters and assisting with questions. * Check-in. Before voters can gain access to a voting booth, poll workers determine their eligibility to vote by verifying their registration using voter lists or poll books--a list of individuals eligible to vote within the voting precinct. Jurisdictions use either paper or electronic poll books--most often laptops or tablets--to check in voters. If the individual does not appear in the poll book for the precinct, federal law requires that an individual asserting to be registered in the jurisdiction for which he or she desires to vote be provided a provisional ballot.[Footnote 25] If individuals are determined to be eligible voters, their provisional ballots are to be counted as votes in accordance with state law, along with other types of ballots, and included in the total election results. * Marking and submitting ballot. After voters are checked in, poll workers direct them to a voting booth to mark their ballots and then submit the ballots for counting. The manner in which votes are cast and counted can vary depending on the voting method and technology employed by the jurisdiction. Currently, most votes are cast and counted by one of two types of electronic voting systems: direct recording electronic (DRE) systems and optical or digital scan systems.[Footnote 26] Such systems include the hardware and software used to define ballots, cast and count votes, report or display election results, and maintain and produce a printed record of voters' selections. Figure 2 shows images for a DRE machine and optical scanners. * DRE machines.[Footnote 27] Voters mark ballots electronically using a touch screen or push-button interface, and their ballot selections are stored in the machine's memory. * Optical or digital scanner. An optical scan system consists of computer-readable paper ballots, appropriate marking devices, privacy booths, and a computerized tabulation device. Optical scan ballots are marked using an appropriate writing instrument to fill in boxes or ovals next to a candidate's name or an issue. If ballots are counted at a central location using a central count optical scan, voters deposit their ballots in a sealed box. If ballots are counted at the polling place using a precinct count optical scan, voters or election officials feed ballots into the scanner. Figure 2: Direct Recording Electronic Machine, Central Count Optical Scanner, and Precinct Count Optical Scanner: [Refer to PDF for image: 3 photographs] Source: GAO. GAO-14-850. [End of figure] Postelection Activities: Following the close of the polls on Election Day, election officials and poll workers complete steps such as securing equipment and ballots, transferring physical ballots or records of precinct vote counts to a central location for counting, and determining the outcome of the election. Votes counted include those cast on Election Day, absentee ballots, early votes (where applicable), and provisional ballots. While preliminary results are available usually by the evening of Election Day, the certified results are generally not available until days later. Election Administration Research and Data Sources on Voter Wait Times: Various studies and research have been conducted on voter wait times. In general, these studies and research have used voter surveys, data on voter check-in and polling place closing times in individual jurisdictions, and the experiences of local jurisdictions to examine issues of voter wait times. For example, the Presidential Commission on Election Administration (Presidential Commission)--established by executive order in March 2013--issued a report in January 2014 identifying best practices in election administration and making recommendations to improve the voting experience and ensure that all eligible voters have the opportunity to cast their ballots without undue delay.[Footnote 28] In conducting its work, the Presidential Commission held public hearings that included academic researchers in the field of election administration and state and local election officials, among others, and surveyed local election officials nationwide regarding a number of issues, including voter wait times. [Footnote 29] Moreover, various researchers have used postelection voter survey data to examine wait time issues, as there is currently no comprehensive set of data that tracks Election Day wait times across precincts nationwide. Two nationwide postelection public opinion surveys in particular have included questions on wait times experienced by voters. The Survey of the Performance of American Elections (SPAE), conducted in 2008 and 2012, is an Internet survey of 200 registered voters in each of the 50 states and the District of Columbia.[Footnote 30] Respondents were asked about their experiences voting--in person on Election Day, in person prior to Election Day, and absentee-- including, for those who reported voting in person, the length of time they recalled waiting in line to vote. The CCES, an Internet survey of U.S. citizens aged 18 and over, has been administered since 2006.[Footnote 31] Researchers have used data from these surveys to estimate individual voter wait times by state, the effect of wait times on the voter experience, and the relationship between demographic characteristics and wait times, among other things. In addition, some researchers have conducted postelection studies of wait time issues for specific states. For example, researchers have assessed the extent of long voter wait times in Maryland, Florida, and other states using data such as voter check-in times, polling place closing times, and information from voter surveys. On the Basis of Our Survey, We Estimate That Most Jurisdictions Did Not Collect Data for Calculating Wait Times or Have Long Voter Wait Times on Election Day 2012: Estimates from our nationwide survey of local election jurisdictions indicate that most jurisdictions did not collect data that would allow them to calculate voter wait times at individual polling places on the November 2012 General Election Day. Our survey found that jurisdictions did not collect these data primarily because wait times have not been an issue. However, some jurisdictions nationwide did collect selected types of wait time data that election officials and researchers have identified would be helpful in measuring wait times. Officials in the jurisdictions we selected for interviews and researchers have measured wait times using various practices, such as the length of time polling places remained open after designated closing times. In addition, estimates from our survey indicate that a small percentage of jurisdictions nationwide had long voter wait times at more than a few polling places on Election Day 2012. Most Jurisdictions Did Not Collect Data for Calculating Wait Times, Primarily because Wait Times Have Not Been an Issue, but Some Have Made Estimates Using Various Practices: On the basis of our survey, we estimate that 78 percent (from 74 to 83 percent) of local jurisdictions nationwide did not collect, receive, or have available information that would allow them to calculate voter wait times that occurred at individual polling places on Election Day 2012.[Footnote 32] Of these jurisdictions, we estimate that 79 percent (from 73 to 84 percent) did not collect this information because, as discussed later in this report, voter wait times have not been an issue.[Footnote 33] Some jurisdictions did collect some types of wait-time-related information that election administration researchers and state and local officials have said could be helpful in measuring wait times. On the basis of our survey, we estimate that officials in jurisdictions nationwide most commonly collected, received, or had available the types of information for Election Day 2012 shown in table 1, such as the number of votes cast at a precinct during a specified time period. These jurisdictions may not have collected these data across all individual polling places. In addition, most of these data may need to be used together or with other information to measure wait times. We estimate that data collection, where it did occur, varied across jurisdictions, with large jurisdictions more likely than small jurisdictions to collect these data. For example, 43 percent of large jurisdictions--those with populations greater than 500,000--collected data on voter complaints about wait times, while 14 percent (from 9 to 19 percent) of small jurisdictions--those with populations between 10,001 and 100,000--collected the same type of data. Table 1: Selected Wait Time Data for Election Day 2012 Collected by Jurisdictions Nationwide: Type of data collected: Observations by election officials of voter wait times at polling places; Estimated percentage of jurisdictions that collected these data: 36 (from 30 to 42 percent). Type of data collected: The number of votes cast at a precinct during a specific time period; Estimated percentage of jurisdictions that collected these data: 31 (from 25 to 36 percent). Type of data collected: The length of time polling places remained open after designated closing times; Estimated percentage of jurisdictions that collected these data: 18 (from 14 to 23 percent). Type of data collected: The time individuals checked into a polling place, recorded by an electronic poll book; Estimated percentage of jurisdictions that collected these data: 17 (from 13 to 22 percent). Type of data collected: Voter complaints about wait times at polling places; Estimated percentage of jurisdictions that collected these data: 16 (from 12 to 21 percent). Source: GAO 2014 survey of local election jurisdictions. GAO-14-850. [End of table] In addition to surveying local election jurisdictions nationwide, we also discussed data collected and practices used for estimating wait times with officials from the 5 jurisdictions we selected for interviews. Officials from these jurisdictions told us that they have estimated wait times at some or all polling places using the data discussed above and other information. For example, * Officials we interviewed at 4 selected jurisdictions stated that they used the time between when the polls closed and when the last voter cast a ballot to estimate wait times. While this technique does not provide information on wait times for the entire voting period, according to researchers we interviewed and studies we reviewed, it could provide reasonable wait time estimates for a small group of voters.[Footnote 34] * Officials from 1 of the jurisdictions said that they also surveyed polling place supervisors on election night regarding the length of voter wait times. Officials from another jurisdiction said that they used election officials traveling across polling locations and precinct poll worker observations to monitor and estimate wait times. Officials from this jurisdiction told us that if these officials or poll workers report problems with voter wait times, actions are taken to address the issue, such as providing additional poll workers or other resources. * Officials at 2 of the 5 selected jurisdictions said that, since the 2012 general election, they have collected wait time data by distributing time-stamped cards to voters upon arrival. In at least one election, 1 of these jurisdictions distributed time-stamped cards to every 15th voter upon arrival. Poll workers then recorded the time on each card at various stages of the voting process and collected the cards when voting was complete. In the other jurisdiction, officials stated that they began measuring wait times from arrival to check-in in the August 2014 election by distributing cards to voters upon arrival and then collecting those cards at the check-in station, where they recorded the time of check-in in an electronic poll book. Some researchers have noted that the study of voter wait times is relatively new because measures of times spent waiting to vote are still being developed.[Footnote 35] These researchers have used a variety of practices to measure wait times, such as polling place closing times--similar to 4 of our 5 selected jurisdictions--or surveys of voters, and primarily have relied on a single type of data. [Footnote 36] Other researchers have measured wait times using multiple sources of data. For example, Maryland's State Board of Elections commissioned a study of Election Day 2012 wait times in jurisdictions within the state, which used available electronic poll book and DRE systems data combined with historical voter turnout information to estimate voter wait times.[Footnote 37] Specifically, the study used the average time to check in voters (recorded via electronic poll books), the number of votes cast per voting machine at a polling place over a fixed time period, and other variables such as expected voter turnout at different times of the day and ballot length to perform a simulation of wait times at all polling places in the state.[Footnote 38] According to the study's authors, election administrators with access to these types of data could use the simulation results, which provided precinct-level information, to see where there are potential problem areas and move resources accordingly when planning for future elections. However, on the basis of our survey, we found that the technologies to generate these types of data are either not available or not being used in many jurisdictions. For example, from our survey, we estimate that electronic poll books were used in 29 percent (from 24 to 34 percent) of local jurisdictions nationwide for the 2012 general election and DRE systems were used in 51 percent (from 45 to 57 percent) of jurisdictions on Election Day. We Estimate That a Small Percentage of Jurisdictions Had Long Voter Wait Times at More than a Few Polling Places on Election Day 2012, and Wait Times Were Believed to Be the Same or Shorter than Those in 2008: Analysis of Our Survey of Local Jurisdictions: Estimates from our nationwide survey of local election jurisdictions indicate that a small percentage of all jurisdictions had long voter wait times at more than a few polling places or long average wait times across polling places on Election Day 2012. In our survey, we defined wait time as the time from when a voter entered the first line to when he or she began filling out a ballot, and we asked jurisdiction officials to estimate (1) how many polling places had wait times that officials considered to be too long, (2) the average voter wait time for all polling places at three times of day, and (3) how many polling places had wait times of greater than 60 minutes at any time during Election Day.[Footnote 39] Because there is no comprehensive set of data on voter wait times across jurisdictions nationwide and we estimate that most jurisdictions did not collect data on wait times based on our survey of election jurisdictions, we relied on election officials in the jurisdictions we surveyed to estimate wait times for these measures based on their perspectives, data, or other information on wait times.[Footnote 40] Our survey asked jurisdiction officials what amount of time they considered to be too long for voters to wait to begin filling out a ballot.[Footnote 41] On the basis of our survey, we estimate that officials in 24 percent (from 19 to 29 percent) of local jurisdictions nationwide believe that a voter wait time of more than 10 minutes on Election Day is too long, officials in 30 percent (from 24 to 35 percent) of jurisdictions believe that a wait time of more than 20 minutes is too long, and officials in 21 percent (from 17 to 26 percent) of jurisdictions believe that a wait time of more than 30 minutes is too long.[Footnote 42] We then asked jurisdiction officials how many of their polling places had wait times that they considered too long on Election Day 2012. The results are shown in figure 3. On the basis of election officials' survey responses, we estimate that 78 percent (from 73 to 83 percent) of jurisdictions nationwide had no polling places with wait times they considered too long on Election Day 2012, 19 percent (from 15 to 23 percent) had a few polling places, and 3 percent (from 1 to 5 percent) had more than a few polling places.[Footnote 43] Figure 3: Estimated Percentage of Local Jurisdictions Nationwide That Had Polling Places with Wait Times Officials Considered Too Long on Election Day 2012: [Refer to PDF for image: vertical bar graph] Percentage of jurisdictions" Number of polling places in jurisdiction: None; 78%; Number of polling places in jurisdiction: A Few; 19%; Number of polling places in jurisdiction: Less than half, but more than a few; 2%; Number of polling places in jurisdiction: About half; 0; Number of polling places in jurisdiction: More than a half; 1%; Number of polling places in jurisdiction: Don't know/no answer; 0. Also depicted: 95% confidence levels. Source: GAO 2014 survey of local election jurisdictions. GAO-14-850. Note: Wait times may be based on officials' perspectives, data, or other information on wait times. Percentages may not add to 100 percent because of rounding. [End of figure] We also asked jurisdiction officials nationwide to estimate the average voter wait time for all polling places at three times of day on Election Day 2012.[Footnote 44] The results are shown in figure 4. On the basis of election officials' survey responses, we estimate that the percentage of jurisdictions with wait times of 0 to 10 minutes during the first hour after the polls opened, around lunchtime, and during the last hour before the polls closed were 51 percent (from 45 to 56 percent), 49 percent (from 43 to 55 percent), and 45 percent, (from 39 to 51 percent) respectively. In addition, we estimate that the percentage of jurisdictions nationwide with wait times of over 20 minutes ranged from 5 percent (from 3 to 8 percent) around lunchtime to 8 percent (from 5 to 12 percent) during the first hour after the polls opened and the last hour before the polls closed.[Footnote 45] According to our survey, in about a third of jurisdictions, officials did not know average voter wait times for these times of day. Figure 4: Estimated Percentage of Local Jurisdictions' Average Wait Times at Polling Places at Different Times on Election Day 2012: [Refer to PDF for image: vertical bar graph] Percentage of jurisdictions: Duration of wait time: O to 10 minutes; First hour after polls opened: 51%; Around lunchtime: 49%; Last hour before polls closed: 45%. Duration of wait time: 11 to 20 minutes; First hour after polls opened: 9%; Around lunchtime: 12%; Last hour before polls closed: 14%. Duration of wait time: Over 20 minutes; First hour after polls opened: 8%; Around lunchtime: 5%; Last hour before polls closed: 8%. Duration of wait time: Don't know/no response; First hour after polls opened: 33%; Around lunchtime: 34%; Last hour before polls closed: 33%. Also depicted: 95% confidence levels. Source: GAO 2014 survey of local election jurisdictions. GAO-14-850. Note: Wait times may be based on officials' perspectives, data, or other information on wait times. Percentages may not add to 100 percent because of rounding. [End of figure] In addition, we asked jurisdiction officials nationwide to estimate how many polling places had wait times of greater than 60 minutes at any time on Election Day 2012. As shown in figure 5, on the basis of officials' survey responses, we estimate that 79 percent (from 75 to 84 percent) of local jurisdictions nationwide had no polling places, 9 percent (from 6 to 13 percent) had a few polling places, and 3 percent (from 1 to 5 percent) had more than a few polling places with wait times of greater than 60 minutes on Election Day 2012. Figure 5: Estimated Percentage of Local Jurisdictions Nationwide That Had Polling Places with Wait Times of Greater than 60 Minutes at Any Time on Election Day 2012: [Refer to PDF for image: vertical bar graph] Number of polling places in jurisdiction: None; Percentage of jurisdictions: 80%. Number of polling places in jurisdiction: A Few; Percentage of jurisdictions: 9%. Number of polling places in jurisdiction: Less than half, but more than a few; Percentage of jurisdictions: 1%. Number of polling places in jurisdiction: About half; Percentage of jurisdictions: 1%. Number of polling places in jurisdiction: More than a half; Percentage of jurisdictions: 1%. Number of polling places in jurisdiction: Don't know/no answer; Percentage of jurisdictions: 9%. Also depicted: 95% confidence levels. Source: GAO 2014 survey of local election jurisdictions. GAO-14-850. Note: Wait times may be based on officials' perspectives, data, or other information on wait times. Percentages may not add to 100 percent because of rounding. [End of figure] Among the 338 jurisdictions that responded to our survey, 18 jurisdictions reported having wait times of greater than 60 minutes at more than a few polling places on Election Day 2012. We assessed these 18 jurisdictions to determine the extent to which they shared any common selected demographic characteristics by analyzing demographic data on these jurisdictions from the U.S. Census. We selected the demographic characteristics to include in our analysis based on those identified in our interviews with researchers and the election administration literature we reviewed as potentially affecting voter wait times. This analysis does not indicate that these demographic characteristics caused voter wait times because there could be other reasons why wait times occurred in these jurisdictions. In addition, the common characteristics we identified apply only to the 18 jurisdictions that reported wait times of greater than 60 minutes and cannot be generalized to the broader election jurisdiction population. On the basis of our analysis, we identified that these 18 jurisdictions tended to have larger populations with lower median ages than survey respondents overall and to have higher proportions of residents who are nonwhite and speak English as a second language. For example, * Twelve of the 18 jurisdictions had large populations (greater than 500,000), 5 had medium-sized populations (100,001 to 500,000), and 1 had a small population (10,001 to 100,000). All but 1 of these 18 jurisdictions had population sizes above the median population size of all respondent jurisdictions. * The median age of the populations of respondent jurisdictions was 38.4 years. Fifteen of the 18 jurisdictions reporting voter wait times of greater than 60 minutes at more than a few polling locations had median ages that were lower than this median. * The median percentage of white residents for the respondent jurisdictions in our sample was 76.4 percent, and 17 of the 18 jurisdictions reporting wait times of greater than 60 minutes at more than a few polling places had white populations below that level, with the lowest being less than 8 percent white. In addition, with regard to the primary language spoken by residents in the jurisdiction, 11 of the 18 jurisdictions reporting long wait times had populations where English is a second language for more than 20 percent of the population, with the highest being 48 percent. The median value for all respondent jurisdictions was a little over 10 percent of residents who spoke English as a second language. Last, we asked jurisdiction officials nationwide to compare typical voter wait times in 2012 and 2008. Our survey results indicate that, according to jurisdiction officials, typical voter wait times for the majority of jurisdictions nationwide on Election Day 2012 were not longer than typical voter wait times on Election Day 2008. We estimate that officials in 44 percent of jurisdictions (from 38 to 50 percent) believed typical wait times were the same in 2012 as 2008, 19 percent (from 14 to 24 percent) believed wait times were shorter in 2012, and 6 percent (from 3 to 9 percent) believed wait times were longer in 2012. In addition, we estimate that in 31 percent (from 26 to 37 percent) of jurisdictions nationwide, officials did not know or were unsure about how wait times in 2012 compared with those in 2008. Our Analysis of Cooperative Congressional Election Study Data and Review of Other Wait Times Studies: In addition to analyzing the results of our survey of a nationally representative stratified random sample of local election jurisdictions, we analyzed voter survey data categorized by state on estimated voter wait times from the CCES.[Footnote 46] Data from the CCES are segmented by state to allow for the comparison of individual voter wait times among states. Our analysis of CCES data demonstrates that, as with the results from our survey of local jurisdictions, long voter wait times were limited on Election Day 2012. Specifically, our analysis of CCES data found that average voter wait times on Election Day 2012 varied across the nation, but few states had average voter wait times of more than 20 minutes. On the basis of this analysis, we estimate that average voter wait times ranged from 1.4 minutes (from 0.7 to 2.1 minutes) in Alaska to more than 34 minutes (from 25 to 43 minutes) in Florida.[Footnote 47] Further, we estimate that in 3 states--Florida, Maryland, and Virginia--about 12 percent or more of voters waited 61 minutes or more to vote.[Footnote 48] Appendix III provides more detailed information on the results of our analysis of CCES data on voter wait times in states. In addition, some studies we reviewed that measured wait times at national and state levels reported that national and statewide average wait times in 2012 generally were not longer than those reported in 2008, and may have been shorter in many cases. For example, one study that used data from two nationally representative surveys of voters indicated that the average wait times for early and Election Day voting combined were lower in 2012 than in 2008 both nationally and in many states.[Footnote 49] Furthermore, this study estimated that the percentage of respondents reporting wait times of less than 10 minutes increased, while the percentage of voters reporting wait times of greater than 60 minutes decreased over the same period. Similarly, in a study conducted for the Maryland State Board of Elections following the 2012 general election, the study's authors found that Maryland residents who reported voting in previous presidential elections and 2012 in a statewide survey were more likely to say that it took less time to vote in 2012 than in earlier elections.[Footnote 50] A Number of Factors Affected Voter Wait Times on Election Day 2012, and Their Impacts Varied across Jurisdictions: On the basis of our survey of local election jurisdictions, interviews with election officials and researchers, and review of relevant literature, we found that various factors, such as voting before Election Day and ballot characteristics, affected voter wait times at different stages in the voting process on Election Day 2012. These factors interacted to affect wait times in the five jurisdictions we selected for interviews, and their impacts varied depending on the unique circumstances in each of the jurisdictions. This variation resulted in targeted approaches by these jurisdictions for reducing wait times where needed and where resources allowed. Various Factors Affected Voter Wait Times: A combination of factors generally affected wait times on Election Day 2012, and these factors may interact to create unique effects on wait times within a jurisdiction or polling place. For instance, one jurisdiction or polling place may be able to manage lines caused by long ballots by increasing the number of voting stations (booths or machines), whereas another jurisdiction or polling place may be unable to set up additional voting stations because it does not have any in reserve, or because there is not enough room at one or more polling places. As a result, the presence of a factor that could contribute to wait times does not necessarily mean wait times will occur. Further, some factors may be present in one election but not another, such as new types of voting equipment, and some factors may be outside the control of local jurisdictions, such as state laws allowing or limiting early in-person voting. It is useful to consider the causes of and solutions for long wait times across jurisdictions to identify common factors; however, it is also important to consider causes and solutions within the unique circumstances of each jurisdiction. The Presidential Commission reported that election administration problems overlap and intersect, and literature we reviewed and researchers we spoke with noted that multiple factors may contribute to long lines, depending on the circumstances of local jurisdictions. As discussed earlier, on the basis of officials' responses to our survey, most local jurisdictions nationwide did not experience long voter wait times on Election Day 2012, and we primarily focused on those that did in assessing the factors involved. For example, we estimate that 22 percent (from 17 to 27 percent) of all jurisdictions nationwide had wait times that officials considered too long at a few or more polling places on Election Day, and we asked officials in these jurisdictions to select which factors they believed contributed to long voter wait times at polling places in their jurisdiction.[Footnote 51] The studies we reviewed have also focused on the effect of long wait times on voters in the 2012 and previous elections. For example, studies found that some individuals were deterred from voting in 2012 and 2008 because of long wait times.[Footnote 52] On the basis of our survey, interviews with state and local election officials and election researchers, and review of literature related to voter wait times, we identified nine key factors that affected wait times on Election Day 2012: * opportunities for voting before Election Day, * type of poll books, * determining voter eligibility, * ballot characteristics, * amount and type of voting equipment, * number and layout of polling places, * number and training of poll workers, * voter education, and: * resource availability and allocation.[Footnote 53] These factors can affect voter wait times at different stages in the voting process on Election Day--(1) arrival, (2) check-in, and (3) marking and submitting the ballot. In addition, some of these factors, such as resource availability and allocation, can cut across multiple stages in the process. Figure 6 shows the voting process and factors that we identified. Figure 6: Voting Stages and Nine Key Factors That Affected Voter Wait Times on Election Day 2012: [Refer to PDF for image: illustration] Stage: Arrival; Factors at each stage: * Opportunities for voting before Election Day. Stage: Check-in; Factors at each stage: * Type of poll books[A]; * Determining voter eligibility. Stage: Mark and submit; Factors at each stage: * Ballot characteristics; * Amount and type of voting equipment. Crosscutting factors: * Number and layout of polling places; * Number and training of poll workers; * Voter education; * Resource availability and allocation. Source: GAO 2014 survey of local election jurisdictions and analysis of studies on election administration and interviews with election officials and researchers. GAO-14-850. [A] A poll book is a list of registered voters and is used by poll workers to verify voters' registration. [End of figure] A key factor that could affect wait times at the arrival stage is: * Opportunities for voting before Election Day. The availability of opportunities to vote before Election Day, such as in-person early voting and mail-in absentee voting, may have affected voter turnout on Election Day 2012. Twenty-seven states and the District of Columbia had laws in effect for the November 2012 election to allow voters to cast an absentee ballot by mail without an excuse. These states and the District of Columbia--as well as 6 additional states--also had laws providing for early voting. Text box: Perspectives on Voting before Election Day: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 24 percent (from 15 to 37 percent), officials believe no or limited opportunities for voting outside of Election Day was a contributing factor; * Election officials in 23 states reported that the availability of alternative voting options, such as voting by mail or early voting can affect wait times; * Of the jurisdictions nationwide making changes to address the causes of long wait times, we estimate that 30 percent (from 18 to 45 percent) are revising polices or procedures related to options for voting outside of Election Day. Source: GAO 2014 survey of local election jurisdictions and analysis of interviews with state election officials. [End of text box] Key factors at the check-in stage that could affect wait times include: * Type of poll books. A poll book is a list of eligible voters assigned to a jurisdiction and is commonly organized alphabetically or by the address of the voters. Poll workers use poll books, whether paper or electronic, at polling place check-in stations to ensure voters are registered, eligible to vote, and at the correct voting precinct. The extent to which a poll book is easily and quickly searched affects the poll worker's ability to efficiently check in voters. On the basis of our national survey of local jurisdictions, we estimate that 29 percent (from 24 to 34 percent) of jurisdictions used electronic poll books and 77 percent (from 72 to 82 percent) used paper poll books on Election Day 2012.[Footnote 54] Text box: Perspectives on Poll Books: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 35 percent (from 26 to 47 percent), officials believe the use of paper poll books was a contributing factor, and in 15 percent (from 7 to 26 percent), officials believe the use of electronic poll books was a contributing factor; * Election officials in 10 states reported that either the state or jurisdictions within the state used or planned to use electronic poll books in an effort to minimize wait times; * The studies we reviewed and researchers we interviewed provided varying perspectives on electronic poll books. For example, - According to a report by the Presidential Commission, electronic poll books can provide several benefits, such as the ability to search for voter information using a variety of fields; - Some researchers noted that electronic poll books may not help with voter wait times or may contribute to them. In particular, one researcher we spoke with said that poll books with limited search capabilities may cause delays in finding voters' registration information. Another researcher noted that older poll workers may not be comfortable with or proficient using the technology. Source: GAO 2014 survey of local election jurisdictions, analysis of interviews with state election officials and election researchers, and Presidential Commission 2014 report on election administration. [End of text box] * Determining voter eligibility. Poll workers must spend additional time determining voter eligibility if the information they are provided at check-in does not match the information in the poll book. This could be due to first-time voters, voters with inactive registration status, voters going to the wrong polling place, or inaccurate voter registration information, among other things.[Footnote 55] First-time voters may be more likely to arrive at the wrong polling place or be unfamiliar with the check-in process. Voters may also arrive at the wrong polling place because redistricting or precinct consolidations led to changes to their polling place locations from previous years.[Footnote 56] Issues with determining voter eligibility can lengthen the transaction time at check-in because poll workers may need to investigate the source of the problem and provide additional assistance, such as administering a provisional ballot.[Footnote 57] Text box: Perspectives on Determining Voter Eligibility: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 44 percent (from 32 to 56 percent), officials believe processing provisional voters was a contributing factor; * Some studies we reviewed noted that the provisional ballot process resulted in longer wait times on Election Day 2012. For example, one study found that voters who cast provisional ballots were at the check- in table twice as long as voters using traditional ballots[A]; * Of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that: - In 49 percent (from 37 to 61 percent), officials believe a large number of first-time voters was a contributing factor; - In 35 percent (from 24 to 48 percent), officials believe that a large number of inactive voters was a contributing factor; - In 35 percent (from 24 to 46 percent), officials believe redistricting and in 28 percent (from 18 to 40 percent), officials believe that consolidation or changes to polling places were contributing factors; - In 24 percent (from 14 to 36 percent), officials believe incorrect or inaccurate voter registration information was a contributing factor; * Researchers and jurisdiction officials we spoke with said that modernizing voter registration would reduce the potential for delays at the polling place. Suggestions include using electronic registration and allowing voters to register and change information online. Source: GAO 2014 survey of local election jurisdictions, analysis of interviews with election researchers and local election jurisdiction officials, and studies on voter wait times. [A] Douglas M. Spencer and Zachary S. Markovits, "Long Lines at Polling Stations? Observations from an Election Day Field Study," Election Law Journal, vol. 9, no. 1 (2010). [End of text box] Key factors at the mark and submit stage that could affect wait times include: * Ballot characteristics. Ballot characteristics, such as their length and design, vary across jurisdictions. These characteristics are subject to state and federal requirements, such as the minority language provisions of the Voting Rights Act and the type of the voting equipment used.[Footnote 58] Text box: Perspectives on Ballot Characteristics: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 71 percent (from 59 to 82 percent), officials believe a long ballot was a contributing factor; * Officials in 15 of the 18 jurisdictions in our survey that reported wait times of greater than 60 minutes at more than a few polling places believed a long ballot was a contributing factor; * Election officials in 16 states and the District of Columbia; * reported that ballot characteristics could affect wait times. Source: GAO 2014 survey of local election jurisdictions and analysis of interviews with state election officials. [End of text box] * Amount and type of voting equipment. The amount of voting equipment allocated to polling places can depend on the level of funding available to jurisdictions to purchase equipment or replace or repair broken equipment and resource planning by jurisdictions.[Footnote 59] Further, the type of voting equipment can affect how efficiently voters mark and submit their ballots, among other things. On the basis of our survey, we estimate that 73 percent (from 68 to 78 percent) of jurisdictions used paper ballots with optical/digital scan counting devices, 51 percent (from 45 to 57 percent) used DRE machines, and 18 percent (from 13 to 24 percent) used paper ballots that were hand- counted on Election Day. Some jurisdictions combined methods. Text box: Perspectives on Amount and Type of Voting Equipment: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that: - In 27 percent (from 17 to 39 percent), officials believe not having enough voting machines was a contributing factor; - In 11 percent (from; * 5 to 20 percent), officials believe the type of voting method or machine used was a contributing factor; * Of the jurisdictions nationwide making changes to address the causes of long wait times, we estimate that: = Twenty-nine percent (from 17 to 44 percent) have revised or are revising polices or procedures related to the number of voting machines; - Eighteen percent (from 9 to 31 percent) have revised or are revising policies or procedures related to the type of voting method or machine used; * Officials in 18 states said that they used policies associated with voting equipment on Election Day 2012 to minimize wait times. In particular, officials in 5 states noted their decision to use optical scan machines to address wait times. According to officials in 1 of these states, using such machines gave them additional flexibility in managing wait times because they could set up additional privacy booths if needed. Source: GAO 2014 survey of local election jurisdictions and analysis of interviews with state election officials. [End of text box] A number of factors could affect voter wait times across more than one stage of the voting process. Such crosscutting factors include: * Number and layout of polling places. Polling places need to meet numerous requirements, including being sizable enough to accommodate the expected number of voters; having sufficient parking available; and complying with federal and state accessibility requirements, including those in the Americans with Disabilities Act.[Footnote 60] In addition, polling places can have differing layouts for moving the voter from arrival to ballot submission. Text box: Perspectives on Number and Layout of Polling Places: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 14 percent (from 6 to 24 percent), officials believe not having enough polling places was a contributing factor, and in 33 percent (from 22 to 45 percent), officials believe the design or layout of polling places was a contributing factor; * Of the jurisdictions making changes to address the causes of long wait times, we estimate that 65 percent (from; * 50 to 78) have revised or are revising policies or procedures related to the design or layout of polling places; * Election officials in 13 states and the District of Columbia reported that issues associated with the polling place, such as location, size, and layout, contributed or could contribute to longer than expected wait times on Election Day; * Some researchers have suggested that schools be used as polling places. For example, the Presidential Commission recommended that states authorize the use of schools as polling places because they typically are large, conveniently located, and comply with federal accessibility requirements. The commission stated that security concerns could be addressed by making Election Day an in-service day for students and teachers. Source: GAO 2014 survey of local election jurisdictions, analysis of interviews with state election officials and election researchers, and Presidential Commission 2014 report on election administration. [End of text box] * Number and training of poll workers. Effective polling place management requires having a sufficient number of poll workers to serve voters and training these workers to efficiently move voters through the voting process and resolve problems. According to our survey, almost all jurisdictions provided standardized training for poll workers. We estimate that the average training jurisdictions provided for typical first-time poll workers was 2.9 hours (from 2.7 to 3.1 hours), and the average training they provided for typical returning poll workers was 2.6 hours (from 2.4 to 2.7 hours) for Election Day 2012. Text box: Perspectives on Number and Training of Poll Workers: * On the basis of our survey results, of the jurisdictions nationwide that had wait times officials considered too long at a few or more polling places on Election Day, we estimate that in 36 percent (from 25 to 48 percent), officials believe training of poll workers was a contributing factor, and in 26 percent (from 16 to 38 percent), officials believe not enough poll workers was a contributing factor; * Of the jurisdictions making changes to address the causes of long wait times, we estimate that: - Seventy-four percent (from 60 to 86 percent) have revised or are revising policies or procedures related to training of poll workers; - Sixty-seven percent (from 52 to 80 percent) have revised or are revising polices or procedures related to the number of poll workers at polling places; * Election officials in 29 states and the District of Columbia said they used or plan to use policies associated with poll workers to minimize wait times. In particular, officials in 21 states emphasized poll worker training as something they used to minimize wait times on Election Day 2012. Source: GAO 2014 survey of local election jurisdictions and analysis of interviews with state election officials. [End of text box] * Voter education. Voter education encompasses providing voters with the information they need to efficiently navigate the voting process. According to our survey, the most common types of information that jurisdictions provided to educate the public prior to the November 2012 general election were specific polling place location information, sample ballots, and information about options to vote before Election Day. Text box: Perspectives on Voter Education: * Of the 291 jurisdictions that responded to our open-ended survey question regarding what policies and procedures were most important to minimizing wait times, 44 cited practices related to voter education; * Election officials in 15 states and the District of Columbia reported that, in an effort to minimize wait times, either the state or jurisdictions within the state took steps to educate voters by, for example, providing sample ballots, polling place information, or real- time information about wait times on Election Day 2012. Source: GAO 2014 survey of local election jurisdictions and analysis of interviews with state election officials. [End of text box] * Resource availability and allocation. The amount of resources available to jurisdictions and how these resources are allocated relate to the other factors we have identified. For example, resource availability and allocation influence the number of voting machines and poll workers in each polling place, as well as jurisdictions' voter education efforts. Jurisdictions' resource planning efforts can encompass preparing for expected voter turnout and requesting and distributing resources to help reduce voter wait times. According to our survey, most jurisdictions had enough resources on Election Day 2012 and resource shortages were rare. We estimate that in 81 percent (from 77 to 85 percent) of jurisdictions, officials believe they had enough resources to comfortably conduct operations on Election Day; in 17 percent (from 13 to 22 percent), officials believe that resources were tight but Election Day operations were conducted as planned; and in 2 percent (from 1 to 4 percent), officials believe there were resource shortages and some Election Day operations were affected by these shortages. Our survey results also show that most jurisdictions tended to use general types of data, such as the number of registered voters, to inform their resource allocation among polling places, and typically did not use more specific measures, such as the estimated average time needed to check in voters. For example, we estimate that 89 percent (from 84 to 92 percent) of jurisdictions used the number of registered voters in each polling place to inform their resource allocation, and 31 percent (from 25 to 36 percent) used the estimated average time needed to check in voters.[Footnote 61] Text box: Perspectives on Resource Availability and Allocation: * Of the 291 jurisdictions that responded to our open-ended survey question regarding what policies and procedures were most important to minimizing wait times, 168--or over half--indicated areas related to Election Day resources, such as better allocation or increasing the number of poll workers or voting machines. In addition, of the 173 jurisdictions that responded to our open-ended question regarding what the federal government could do to help address long voter wait times, 100 mentioned providing funding or enacting policies related to additional election resources, such as voting equipment; * Studies we reviewed and researchers we interviewed noted a relationship between the allocation of resources and wait times. For example, - The Presidential Commission noted that it is usually the allocation of resources between polling places, rather than the total resources available, that causes long lines. The Presidential Commission and researchers with whom we spoke suggested jurisdictions seeking to reduce voter wait times use resource allocation methods that incorporate targeted inputs--such as estimated turnout by hour and estimated average service times for voter check-in and ballot completion; - The Brennan Center for Justice noted relationships between resource shortages and voter wait times at the end of the day on Election Day 2012 after analyzing relevant data for studied counties in Florida, Maryland, and South Carolina. The Brennan Center reported that for the selected counties in Florida and South Carolina, the 2 states they examined that had data on how poll workers were allocated, lines were generally longer when poll workers had to serve more voters. The Brennan Center also reported that, in general, for studied counties across all 3 states, the more registered voters a machine had to serve, the longer the delay.[A] Source: GAO 2014 survey of local election jurisdictions, Brennan Center 2014 report on Election Day long lines, and Presidential Commission 2014 report on election administration. [A] Brennan Center for Justice, New York University School of Law, Election Day Long Lines: Resource Allocation (New York, New York: September 2014). [End of text box] Impact of Factors Varied across Jurisdictions: Multiple factors affected wait times on Election Day 2012 in the 5 local jurisdictions we selected for interviews. We selected these jurisdictions based on wait times, election administration policies, and demographic characteristics, among other things, to reflect a range of local experiences and illustrate how factors affected wait times in practice. While the nine factors we identified apply generally across all jurisdictions, their specific impact on Election Day 2012 wait times depended on the unique circumstances in each of our selected jurisdictions, leading to targeted approaches for reducing wait times where needed and where resources allowed. Table 2 summarizes reported wait times and election administration characteristics related to these nine factors across our 5 selected jurisdictions. Table 2: Reported Wait Times and Characteristics on Election Day 2012 across Five Selected Jurisdictions: Reported wait times: Reported polling places with wait times of greater than 60 minutes on November 2012 Election Day; Jurisdiction 1: More than half of all polling places; Jurisdiction 2: More than half of all polling places; Jurisdiction 3: About half of all polling places; Jurisdiction 4: Less than half but more than a few polling places; Jurisdiction 5: No polling places. Factor: Reported jurisdiction characteristics: Number of registered voters; Jurisdiction 1: 248,940; Jurisdiction 2: 586,854; Jurisdiction 3: 51,357; Jurisdiction 4: 388,425; Jurisdiction 5: 4,674,338. In-person voting on Election Day; Jurisdiction 1: 155,176; Jurisdiction 2: 210,183; Jurisdiction 3: 33,731; Jurisdiction 4: 126,351; Jurisdiction 5: 2,577,135. Opportunities for voting before Election Day: Percentage of votes cast before Election Day (e.g., any type of early, absentee, and mail-in ballots); Jurisdiction 1: 14%; Jurisdiction 2: 28%; Jurisdiction 3: 8%; Jurisdiction 4: 53%; Jurisdiction 5: 27%. Type of poll books: Type of poll book used; Jurisdiction 1: Electronic; Jurisdiction 2: Electronic and paper; Jurisdiction 3: Paper; Jurisdiction 4: Paper; Jurisdiction 5: Paper. Determining voter eligibility: Percentage provisional voting on Election Day; Jurisdiction 1: 0.2%; Jurisdiction 2: .01%; Jurisdiction 3: 1%; Jurisdiction 4: 1%; Jurisdiction 5: 12%. Ballot characteristics: Average number of pages/screens on ballots; Jurisdiction 1: 6 screens; Jurisdiction 2: 4 pages; Jurisdiction 3: 2 pages; Jurisdiction 4: 8 pages; Jurisdiction 5: 7 pages. Total number of elected offices and ballot questions placed on applicable ballots jurisdiction-wide; Jurisdiction 1: 6 elected offices; 2 ballot questions; Jurisdiction 2: 26 elected offices; 18 ballot questions; Jurisdiction 3: 6 elected offices; 1 ballot question; Jurisdiction 4: 54 elected offices; 14 ballot questions; Jurisdiction 5: 90 elected offices; 53 ballot questions. Ballot in more than one language; Jurisdiction 1: No; Jurisdiction 2: No; Jurisdiction 3: Yes; Jurisdiction 4: Yes; Jurisdiction 5: Yes. Amount and type of voting equipment: Predominant type of voting method; Jurisdiction 1: Direct recording electronic; Jurisdiction 2: Optical scan; Jurisdiction 3: Optical scan; Jurisdiction 4: Optical scan; Jurisdiction 5: Optical scan. Number and layout of polling places: Number of precincts and polling places; Jurisdiction 1: 77 precincts; 77 polling places; Jurisdiction 2: 490 precincts; 200 polling places; Jurisdiction 3: 24 precincts; 24 polling places; Jurisdiction 4: 125 precincts; 88 polling places; Jurisdiction 5: 4,993 precincts; 4,621 polling places. Average number of registered voters per precinct; Jurisdiction 1: 3,233; Jurisdiction 2: 1,198; Jurisdiction 3: 2,140; Jurisdiction 4: 3,107; Jurisdiction 5: 936. Number and training of poll workers: Average number of poll workers per precinct; Jurisdiction 1: 10; Jurisdiction 2: 12; Jurisdiction 3: 16; Jurisdiction 4: 12; Jurisdiction 5: 6. Poll worker training (hours); Jurisdiction 1: Supervisor--3.5; First-time worker--2; Returning worker--2; Jurisdiction 2: Supervisor--10; First-time worker--15; Returning worker--5; Jurisdiction 3: Supervisor--8; First-time worker--8; Returning worker-- 8; Jurisdiction 4: Supervisor--5; First-time worker--4; Returning worker-- 3; Jurisdiction 5: Supervisor--3; First-time worker--2.5; Returning worker--2.5. Voter education: Sample ballots by mail; Jurisdiction 1: No; Jurisdiction 2: Yes; Jurisdiction 3: No; Jurisdiction 4: Yes; Jurisdiction 5: Yes. Specific polling place information by mail; Jurisdiction 1: Yes; Jurisdiction 2: Yes; Jurisdiction 3: No; Jurisdiction 4: Yes; Jurisdiction 5: Yes. Resource availability and allocation: Views on availability of resources; Jurisdiction 1: Resource shortages; Jurisdiction 2: Enough resources; Jurisdiction 3: Resource shortages; Jurisdiction 4: Resource shortages; Jurisdiction 5: Enough resources. Source: GAO 2014 survey of local election jurisdictions. GAO-14-850. [End of table] Jurisdiction 1: Jurisdiction 1 reported in our survey that more than half of all its polling places had wait times of greater than 60 minutes on Election Day 2012. In addition, the last voter checked in over 2 hours after the precinct's designated closing time at 10 of the jurisdiction's 77 precincts, according to data provided by jurisdiction officials. The jurisdiction reported that Election Day wait times in 2008 were about the same as those in 2012, but officials said that the lines in 2008 were concentrated in the morning with few lines at closing. Officials stated that the long lines on Election Day 2012 were caused by a variety of factors. According to jurisdiction officials, an insufficient number of voting machines was a primary cause of long voter wait times. Officials said that state laws (1) prohibited the jurisdiction from purchasing additional DRE voting machines and (2) did not allow for the use of on- hand emergency paper ballots that could be hand-counted in addition to the use of DRE machines.[Footnote 62] As a result, the jurisdiction used roughly the same number of DRE machines in 2012 as in 2008 despite a 14 percent increase in Election Day votes cast, and was unable to deploy paper ballots to help mitigate long lines. Officials stated that they were limited in their ability to add voting capacity at polling places with the longest lines because all DRE machines were already deployed and it was not practical to reprogram the machines for use at precincts that may have had greater needs. Issues with determining voter eligibility, primarily resulting from the large number of inactive voters, was another key factor that contributed to long wait times.[Footnote 63] Voters marked as inactive on the jurisdiction's registration rolls were required to fill out Affirmation of Eligibility forms to verify their addresses, and the number of these forms filled out by voters on Election Day 2012 was more than 12 times higher than in 2008.[Footnote 64] Further, according to the jurisdiction's postelection report, available data indicate that every voter requiring an address confirmation using this process resulted in three additional voters checking in after precinct closing time.[Footnote 65] Jurisdiction officials explained that their policy was to move all voters with check-in problems out of the main line and direct them to the precinct supervisor for resolution. Officials said that this helped to alleviate wait times at the main check-in line to some extent, but there were still delays because of the time it took to identify issues and direct voters to the supervisor, among other things. In addition, officials noted that these voters diverted resources from voters without difficulties. Officials also said that according to an analysis performed by the former Acting Registrar and a postelection task force established by the jurisdiction, the biggest factor in late precinct closing times was the number of voters with issues requiring assistance from the precinct supervisor. In addition to issues with determining voter eligibility, other matters that required assistance from the supervisor included voters with disabilities or voters over the age of 65 who chose to vote from their vehicles outside the polling place (curbside voters), as allowed by state law, and voters who brought their absentee ballots to precincts on Election Day. Officials noted that when precinct supervisors called the registrar's office to resolve issues that required additional assistance, the office could not keep up with demand. The large numbers of voters in each precinct also contributed to long lines. Specifically, according to the jurisdiction's postelection report, there was a high correlation between large precincts and the number of citizens voting after precinct closing time, and two of the four precincts with over 5,000 registered voters checked in their last voter over 2-1/2 hours after closing time. Finally, jurisdiction officials said that recruitment and retention of poll workers was challenging. According to these officials, poll worker experience is important for managing polling locations on Election Day, but there is a roughly 50 percent drop in retention from election to election, and presidential elections generally require a large number of poll workers to handle greater levels of voter turnout. Jurisdiction officials stated that one precinct in particular faced a confluence of factors that contributed to its remaining open nearly 4 hours after its official closing time on Election Day 2012. Text box: Example of How Factors Combined to Create Long Wait Times in an Individual Precinct: Jurisdiction officials and a precinct official cited several factors that combined to create long wait times: * Total turnout increased by 63 percent--from 2,104 in 2008 to 3,425 in 2012; * Large number of first-time and inactive voters who can take longer to check in and vote[A]; * Large number of curbside voters, which under state law required the assistance of two poll workers when a portable electronic voting device was used; * Large number of voters needing language assistance. Such assistance required that both the voter and the assisting poll worker fill out separate forms before the poll worker provided translation services.[B] Source: Analysis of interviews with officials from the local election jurisdiction and a precinct within the jurisdiction. [A] The Acting Registrar on Election Day 2012 stated that the jurisdiction retrieved data on the time spent voting from three voting machines in the jurisdiction, one of which was in this precinct. He said that these data showed that on average, voters in this precinct took 3 minutes and 30 seconds to cast their ballots, whereas voters using the voting machines from the other two precincts took an average of between 1 minute and 1 minute and 20 seconds. [B] For Election Day 2012, this jurisdiction did not meet the requirements of the coverage formula under section 203 or 4(f)(4) of the Voting Rights Act, and as a result, was not required to provide instructions and ballots, among other things, in the language of applicable minority groups, as well as in English. Voting Rights Act Amendments of 2006, Determinations Under Section 203, 76 Fed. Reg. 63602-07 (Oct. 11, 2013); 28 C.F.R. § 55.5. [End of text box] According to officials, jurisdiction 1 has made or is in the process of making several changes to address the causes of long voter wait times on Election Day 2012. The jurisdiction reported that it is replacing its DRE voting machines with optical scanners (for scanning paper ballots), which allow for more flexibility in adding voting station capacity. The jurisdiction also plans on maintaining adequate scanners and personnel in reserve to handle any equipment failures that might arise, and officials noted that scanners can be easily transferred among precincts. In addition, the jurisdiction reported that it is purchasing more electronic poll books to help check in voters more efficiently and that its Electoral Board has tested scanners (capable of scanning driver's licenses and voter cards) to be used with the electronic poll books for instantaneous and accurate voter check-in. Additionally, the jurisdiction has added 14 precincts and reduced the largest precincts to fewer than 4,000 registered voters, according to jurisdiction officials. The jurisdiction's Electoral Board believes that these and other changes will allow it to meet the jurisdiction's newly established goal of having no voters expend more than 30 minutes from the time they arrive at the polling place until they cast their ballots. Jurisdiction 2: Jurisdiction 2 reported in our survey that more than half of its polling places had wait times of greater than 60 minutes on Election Day 2012. In addition, it reported average wait times of more than 120 minutes in the hour after polls opened and the hour before polls closed. Jurisdiction officials stated that despite the long voter wait times in 2012, things went more smoothly than in 2008. This was in part because of measures implemented to address the issue, such as voter education and increasing the number of voting stations at polling places, as advised by state guidance distributed shortly before the 2012 presidential election.[Footnote 66] According to jurisdiction officials, the length of the ballot--four 22- inch pages--was the primary cause of long voter wait times. Officials noted that ballot length, including the number of elected offices and ballot questions, is outside of the jurisdiction's control. Prior to the election, the jurisdiction tested the time required to complete the ballot on a variety of constituencies and found that it took voters unfamiliar with the ballot between 12 and 15 minutes to complete it.[Footnote 67] Officials stated that they implemented measures to try to alleviate the expected congestion at the voting booths, such as increasing the number of booths at polling locations, mailing sample ballots to registered voters, and employing line monitors to help ensure that voters were prepared to vote immediately after check-in.[Footnote 68] Furthermore, officials stated that in some instances, lines formed at the scanning machines because voters were trying to feed the long ballots into the machines too quickly and caused paper jams. In these cases, technicians were available to address problems and poll workers stepped in to take over the scanning. The jurisdiction also maintained a reserve supply of 100 scanners to replace machines that could not be fixed by technicians. Nonetheless, the jurisdiction reported that voters waited in line an average of 16 to 30 minutes to turn in their ballots after completing them. According to officials, the jurisdiction does not plan to make any changes to its policies and procedures to address long wait times. They noted that the long wait times in the 2012 election were primarily caused by a lengthy ballot, which is outside of their control. Jurisdiction 3: Jurisdiction 3 reported in our survey that about half of its polling places had wait times of greater than 60 minutes on Election Day 2012, and that typical wait times in 2012 were less than those in 2008. According to jurisdiction officials, the additional time spent determining voter eligibility, and issuing provisional ballots, as necessary, was the primary cause for long lines in their jurisdiction. Key factors contributing to delays in determining voter eligibility were: * Redistricting. Jurisdiction officials stated that redistricting increased the number of voters who went to the wrong polling location because their polling place had changed from the previous election. These officials explained that redistricting changed the location of all but one of the voting precincts in their jurisdiction.[Footnote 69] * Inactive voters. According to jurisdiction officials, there was a large number of inactive voters who had not voted in at least two general elections. Officials stated that delays were caused by time spent assisting these voters--by, for example, calling the election office to request further research--and updating their status on voter registration lists. * Inaccurate voter registration information. According to jurisdiction officials, inaccurate voter registration information may have also caused delays in determining voter eligibility. An official stated that poll workers directed most voters who said they were registered but were not included in the state's list of registered voters to the jurisdiction's central office, where Election Day registration was conducted.[Footnote 70] An investigation initiated by the state after Election Day found that some voter registration forms had not been forwarded to the election office and a number of voter requests for registration information had not been processed by a third-party state agency.[Footnote 71] Jurisdiction officials believe that this contributed to the number of individuals who thought they had submitted registration materials but were not registered, leading to delays in checking in voters on Election Day. In addition, officials noted that many voters were unaware of the need to update their registration information with each address change. This led to voters arriving at the wrong polling places and requiring additional assistance during check-in. Further, jurisdiction officials stated that redistricting required locating new polling places to serve the new precincts and noted the challenges they faced in doing so. Specifically, they said that it was difficult to find polling places that had the necessary space and parking and were compliant with federal and state polling place accessibility requirements. They noted that they primarily try to use churches, community centers, and schools as polling places. However, they said that while schools tend to have the necessary space and layout to help process voters efficiently, using schools has become more challenging in recent years because of security concerns. In addition, principals are concerned about the potential disruptions that the polling activities might have on students. Jurisdiction officials noted that 9 schools were used as polling places in 2012, down from the 14 generally used during previous elections. Jurisdiction officials said that their elections budget limits the changes they can implement to address wait times. For example, officials stated that additional voter education could help address some of the issues with determining voter eligibility that were experienced on Election Day 2012, such as voters arriving at the wrong polling place, but their jurisdiction lacks the resources to provide this education. Jurisdiction 4: Jurisdiction 4 reported in our survey that less than half, but more than a few, of its polling places had wait times of greater than 60 minutes on Election Day 2012. In addition, 26 percent of its precincts remained open over 3 hours after the designated closing time, according to data provided by jurisdiction election officials. These officials stated that the jurisdiction did not have wait time issues in 2008 of which they were aware and that a confluence of factors created the long wait times in 2012. According to jurisdiction officials, lengthy ballots--an average of eight 17-inch pages--were the primary cause of long voter wait times. Factors contributing to the long ballots included (1) 12 state constitutional amendments that spanned five pages; (2) state requirements to include special district races, such as fire control, mosquito control, and community development districts, on the ballots; [Footnote 72] and (3) the Voting Rights Act requirement to include both English and Spanish on the ballots.[Footnote 73] Officials said that voters took a significant amount of time to fill out their ballots, which resulted in congestion at voting booths in some polling places. In addition, these officials stated that the long ballots led to paper jams when voters scanned their ballots, contributing to average wait times of more than 30 minutes to turn in ballots after completing them. According to jurisdiction officials, the length of the ballot for the November 2012 election was not determined until mid- June because of the timing for determining qualifying candidates and state amendments. Officials said that they sent out sample ballots to help educate voters, but did not have sufficient time to effectively plan or take additional actions to help mitigate the effects. Jurisdiction officials said that the second key cause of long wait times on Election Day was state-wide reductions in the number of days and limited locations available for in-person early voting. State legislation changed the number of allowable early voting days from 14 days in 2008 to 8 days in 2012.[Footnote 74] In addition, officials said that existing state law limited the sites that could be used for early voting to a few types of public buildings, which created challenges with finding locations that could conveniently and effectively serve voters.[Footnote 75] Election officials reported that while there were about 5,500 fewer total voters in 2012 compared with the 2008 general election (a 2 percent decrease), approximately 14,500 fewer people voted early (a 22 percent decrease). As a result, more voters than expected came to the polls on Election Day and the jurisdiction did not have enough resources to effectively accommodate them, according to officials. For example, officials stated that there was an insufficient number of voting booths and scanners in some polling locations because of both the larger than expected turnout and the time needed to fill out and scan long ballots. Jurisdiction officials said that other factors also contributed to wait times. For example, they stated that redistricting and precinct consolidations may have increased the number of voters who went to the wrong polling location because their polling place had changed from the previous election.[Footnote 76] Officials stated that they consolidated precincts--reducing the number from 171 in 2008 to 125 in 2012--in an effort to lower election expenditures, but that this may have contributed to long lines by increasing the number of voters in certain polling locations. Further, officials said that heavy rain on Election Day led to line management issues in some locations because poll workers were checking in more voters than their polling place could accommodate to get voters out of the rain. This contributed to the congestion at the voting booths and scanning machines. Jurisdiction officials said that a number of changes have been made since the 2012 election to address long wait times. For example, state laws established additional word limits to state constitutional amendments on the ballot,[Footnote 77] restored the allowable number of early voting days to 2008 levels, and expanded the types of sites that can be used for early voting locations.[Footnote 78] In addition, the jurisdiction has replaced the paper poll books used in 2012 with electronic poll books, which officials anticipate will help expedite the check-in process; better estimate voter turnout for resource allocation; and allow them to post current wait times for each precinct online, which would help voters identify times to go to the polls if they do not want to wait. According to officials, the jurisdiction also purchased an additional 100 optical scanners so that each precinct will have 2 scanners. In addition, officials said that they have launched voter education efforts, such as public service announcements on radio, television, and other forms of media, to encourage mail-in voting and inform voters about how to access voter registration and polling place information. Further, these officials noted that as of 2014, the jurisdiction began paying the return postage on mail-in ballots. Jurisdiction 5: Jurisdiction 5 reported in our survey that it had no polling places with wait times of greater than 60 minutes on Election Day 2012 and that typical wait times in 2012 were less than those in 2008. This jurisdiction had ballots that were seven pages long on average and a high percentage of provisional voters but reported that it did not experience long voter wait times. Jurisdiction officials said that the county provided sufficient resources for conducting the 2012 general election, which helped ensure that long wait times did not occur. According to officials, this allowed election planners to include a safety margin when allocating resources in case of a larger than expected turnout and deploy large amounts of additional resources to polling places that were expected to have higher numbers of voters or substantial issues with determining voter eligibility on Election Day 2012. Jurisdiction officials reported that provisional voting accounted for 12 percent of total in-person voting on Election Day 2012.[Footnote 79] These officials said that the jurisdiction's practice of taking individuals whose eligibility to vote is unclear out of the main check- in line and administering provisional ballots, if needed, in another area of the polling place was important to reducing wait times. They also noted that the jurisdiction's policy of permitting all voters experiencing eligibility issues to vote provisionally and not adjudicating these issues at the polling place reduced the time and resources expended on problems at check-in. In addition, state law allowed individuals to vote provisionally in precincts other than the one to which they were assigned and have applicable votes on their ballots counted.[Footnote 80] This helped expedite the processing of individuals whose eligibility to vote was unclear and reduced wait times, according to jurisdiction officials. Jurisdiction officials stated that a number of other policies and practices helped minimize wait times on Election Day 2012. For example, * Officials said that limiting the number of voters in each precinct helped prevent overcrowding and congestion at polling places. State law mandates a maximum of 1,000 registered voters per precinct.[Footnote 81] Officials noted that they need to ensure the jurisdiction has sufficient polling places, poll workers, and voting equipment to support the number of precincts required to meet this requirement. * According to jurisdiction officials, mailing sample ballots to registered voters helped to shorten the time it took to fill out ballots in the voting booth.[Footnote 82] A polling place inspector-- who supervises polling place operations and staff--we interviewed said that the lengthy ballot did not result in long wait times at her polling place on Election Day 2012 because many voters brought in their sample ballots and knew how they would vote. She also noted that this helped facilitate check-in because voters' names and addresses were on the sample ballot. * Officials said that permanent absentee voting and no-excuse absentee voting by mail, permitted by state law, reduced the number of voters on Election Day. * Officials also stated that poll worker training, which includes how to assist provisional voters and what to do if wait-time-related issues arise, and having experienced poll workers were important to ensuring minimal lines at polling places.[Footnote 83] According to officials, the majority of poll workers have served in previous elections. In addition, officials said that having a pool of reserve poll workers who can fill in at locations that need additional staff was essential to reducing wait times. * The jurisdiction deployed mobile units that travel to polling places and distribute additional ballots and other supplies if needed, according to officials. Agency and Third-Party Comments: We provided a draft of this report to the EAC for review and comment. The EAC had no comments on the draft report, as noted in an e-mail received on September 17, 2014, from the commission's Acting Executive Director. We also provided excerpts of the draft report to the chief election officials of each of the 5 local election jurisdictions that we selected for interviews. The excerpts for each of these jurisdictions included findings that pertained specifically to the individual jurisdiction and a description of the methodology we employed to select the 5 jurisdictions. One jurisdiction provided written comments on the excerpts provided for review, which are reproduced in full in appendix IV. The chief election official from this jurisdiction stated that our description of the jurisdiction's experiences was accurate, noted that a series of issues contributed to the long wait times in the jurisdiction on Election Day 2012, and noted the actions the jurisdiction had taken to address them. One jurisdiction provided technical comments, which we incorporated in the report as appropriate. Three jurisdictions reviewed the excerpts and indicated that they had no comments in e-mails received from the jurisdictions' chief election officials on September 8, September 17, and September 22, 2014. We are sending copies of this report to the Election Assistance Commission, election offices in the 5 selected local jurisdictions that participated in our research, appropriate congressional committees and members, and other interested parties. In addition, this report is available at no charge on GAO's website at [hyperlink, http://www.gao.gov]. If you or your staff have any questions, please contact Rebecca Gambler at (202) 512-8777 or gamblerr@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made significant contributions to this report are listed in appendix V. Signed by: Rebecca Gambler: Director, Homeland Security and Justice: [End of section] Appendix I: Objectives, Scope, and Methodology: This report addresses the following questions: 1. To what extent did local election jurisdictions collect data to measure voter wait times and have long voter wait times on Election Day 2012? 2. What factors affected voter wait times on Election Day 2012, and what were the impacts of these factors across jurisdictions? For both objectives, we (1) conducted a web-based survey of election officials from a nationally representative stratified random sample of 423 local election jurisdictions, excluding jurisdictions with populations of 10,000 or fewer and jurisdictions in Oregon and Washington;[Footnote 84] (2) analyzed responses from the 2012 Cooperative Congressional Election Study (CCES), a survey of U.S. citizens aged 18 and over;[Footnote 85] (3) interviewed state election officials from 47 states and the District of Columbia,[Footnote 86] as well as local election administration officials, on-site or by phone, from 5 selected local jurisdictions--Detroit, Michigan; Hartford, Connecticut; Lee County, Florida; Los Angeles County, California; and Prince William County, Virginia; and (4) interviewed officials from the Election Assistance Commission (EAC) and 14 researchers and representatives from research organizations in the field of election administration. We also reviewed relevant literature on voter wait times, such as studies on wait times by various researchers and reports completed or sponsored by state or local governments in our 5 selected jurisdictions.[Footnote 87] A GAO social scientist and a GAO statistician reviewed the studies whose findings we cite in this report and determined that the design, implementation, and analyses of the studies were sufficiently sound to support the studies' results and conclusions based on generally accepted social science principles. 2014 Survey of Local Election Jurisdictions: To obtain national information from local election officials on voter wait times on Election Day 2012, we conducted a web-based survey of election officials from a stratified random sample of 423 local election jurisdictions. We surveyed officials about any data their jurisdictions collected related to wait times on Election Day 2012, voter wait times in their jurisdictions on this day, and their views on factors that affected long voter wait times, among other things. We defined wait time as the time from when a voter entered the first line to when he or she began filling out a ballot. Reported wait times may be based on officials' perspectives, data, or other information on wait times.[Footnote 88] Our survey period was from March 20, 2014, through June 6, 2014, and we received 338 completed surveys for an overall response rate of 80 percent.[Footnote 89] Overall, there are about 10,500 local government jurisdictions responsible for conducting elections nationwide. States can be divided into two groups according to how they delegate election responsibilities to local jurisdictions. The first group is composed of 41 states that delegate election responsibilities primarily to counties, with a few of these states delegating election responsibilities to some cities, and 1 state that delegates these responsibilities to election regions. We included the District of Columbia in this group of states. The first group contains about one- fourth of the local election jurisdictions nationwide. The second group is composed of 9 states that delegate election responsibilities to subcounty governmental units, known by the U.S. Census Bureau as minor civil divisions (MCD). This group of states contains about three- fourths of the local election jurisdictions nationwide. The categorization of the 50 states and the District of Columbia by how election responsibilities are organized is as follows (states in bold delegate election responsibilities to some cities independently from counties): * County-level states: Alabama, Alaska (four election regions), Arizona, Arkansas, California, Colorado, Delaware, the District of Columbia, Florida, Georgia, Hawaii, Idaho, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Mississippi, Missouri, Montana, Nebraska, Nevada, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia, Washington, West Virginia, and Wyoming: * Minor civil division-level states: Connecticut, Maine, Massachusetts, Michigan, Minnesota, New Hampshire, Rhode Island, Vermont, and Wisconsin: While only about one-fourth of election jurisdictions nationwide are in states that delegate election responsibilities primarily to counties, according to the 2010 Census, 88 percent of the U.S. population lived in these states. The U.S. population distribution between the two state groups is shown in table 3. Table 3: Population in Each State Group: State group: County-level states; Population in 2010: 269,700,327; Percentage: 88%. State group: Minor civil division-level states; Population in 2010: 35,319,416; Percentage: 12%. State group: Total; Population in 2010: 305,019,743; Percentage: 100%. Source: GAO analysis of U.S. Census Bureau data. GAO-14-850. [End of table] Our sampling unit was the geographically distinct local election jurisdiction at the county, city, or MCD level of local government (or, in Alaska, the election region). The initial list of jurisdictions for each state group above was constructed from the 2010 decennial Census data. Census population data were available for all counties, county equivalents, and MCDs.[Footnote 90] We excluded the states of Oregon and Washington because, as of the November 2012 general election, they both were vote-by-mail states where individuals generally do not go to polling places to vote. As a result, our sample frame included jurisdictions in 48 states and the District of Columbia. In addition, we excluded about 7,600 jurisdictions with populations of 10,000 or fewer because, on the basis of our review of wait time research, jurisdictions of this size were unlikely to have experienced long voter wait times. We divided each state group--county-level and MCD-level--into strata according to jurisdiction population size. We used jurisdiction population size, rather than the number of eligible or registered voters, to define sample strata because these Census data were readily available for all counties and MCDs nationwide.[Footnote 91] County- level states were divided into six strata, and MCD-level states were divided into five strata. The allocation of units, or jurisdictions, to strata is shown in table 4. We included all 108 jurisdictions in strata with populations of greater than 500,000--strata 1, 2, and 7-- in our sample because, on the basis of our review of wait time research, the largest jurisdictions were most likely to have experienced long voter wait times. We then selected random samples of jurisdictions in each of the remaining strata, applying a minimum allocation of 20 jurisdictions per stratum. This resulted in a total sample of 423 jurisdictions. Our sample allocation also allowed us to have a random sample of local jurisdictions nationwide according to population size--large, medium, and small. To group jurisdictions by population size, we combined jurisdictions in like-sized population strata in county-level and MCD-level states. We defined large jurisdictions as those with a population greater than 500,000 (strata 1, 2, and 7), medium jurisdictions as those with a population of more than 100,000 to 500,000 (strata 3 and 8), and small jurisdictions as those with a population of more than 10,000 to 100,000 (strata 4, 5, 6, 9, 10, and 11). Upon completion of the survey, we adjusted the sampling weights for nonresponse.[Footnote 92] Table 4: Local Jurisdiction Election Survey Sample Allocation: Stratum: County/city--greater than 1,000,000; Number of jurisdictions in population: 34; Number of jurisdictions sampled: 34. Stratum: County/city--from 500,001 to 1,000,000; Number of jurisdictions in population: 71; Number of jurisdictions sampled: 71. Stratum: County/city--from 100,001 to 500,000; Number of jurisdictions in population: 383; Number of jurisdictions sampled: 41. Stratum: County/city--from 50,001 to 100,000; Number of jurisdictions in population: 326; Number of jurisdictions sampled: 35. Stratum: County/city--from 25,001 to 50,000; Number of jurisdictions in population: 543; Number of jurisdictions sampled: 56. Stratum: County/city--from 10,001 to 25,000; Number of jurisdictions in population: 752; Number of jurisdictions sampled: 76. Stratum: Minor civil division--from 500,001 to 1,000,000; Number of jurisdictions in population: 3; Number of jurisdictions sampled: 3. Stratum: Minor civil division--from 100,001 to 500,000; Number of jurisdictions in population: 22; Number of jurisdictions sampled: 20. Stratum: Minor civil division--from 50,001 to 100,000; Number of jurisdictions in population: 79; Number of jurisdictions sampled: 20. Stratum: Minor civil division--from 25,001 to 50,000; Number of jurisdictions in population: 175; Number of jurisdictions sampled: 20. Stratum: Minor civil division--from 10,001 to 25,000; Number of jurisdictions in population: 448; Number of jurisdictions sampled: 47. Stratum: Total; Number of jurisdictions in population: 2,836; Number of jurisdictions sampled: 423. Source: GAO analysis of U.S. Census Bureau data and allocation of jurisdictions to sample. GAO-14-850. [End of table] We analyzed survey responses to provide nationwide estimates for Election Day 2012 of any data collected on wait times, voter wait times, views on the factors that affected long wait times, policies and practices used, and any revisions to policies to address the possible causes of long wait times, among other things.[Footnote 93] All sample surveys are subject to sampling error--that is, the extent to which the survey results differ from what would have been obtained if the whole population had been observed. Because we followed a probability procedure based on random selections, our sample is only one of a large number of samples that we might have drawn. As each sample could have provided different estimates, we express our confidence in the precision of our particular sample's results as a 95 percent confidence interval (e.g., from x to y percent). This is the interval that would contain the actual population value for 95 percent of the samples we could have drawn. As a result, we are 95 percent confident that each of the confidence intervals based on our survey includes the true values in the sample population. In addition to the reported sampling errors, the practical difficulties of conducting any survey may introduce other types of errors, commonly referred to as nonsampling errors. For example, differences in how a particular question is interpreted, the sources of information available to respondents, or the types of people who do not respond can introduce unwanted variability into the survey results. We took numerous steps in questionnaire development, data collection, and the editing and analysis of the survey data to minimize nonsampling errors. For example, a social science survey specialist designed the draft questionnaire for local jurisdictions in close collaboration with GAO subject matter experts. We also utilized information from prior GAO reports, our review of studies on wait times, and interviews with election administration researchers, discussed below, to help inform the development of the questionnaire.[Footnote 94] In addition, we pretested the survey in person or by telephone with officials in 7 election jurisdictions of various sizes in 5 states and made revisions, as necessary. The survey questionnaire and aggregated responses for each question are included in appendix II. Further, we omitted responses on all completed surveys that fell outside of specified limits, such as when the reported number of ballots cast was greater than the reported number of registered voters in a jurisdiction, and called respondents in some cases to obtain information where clarification was needed. Analysis of 2012 Cooperative Congressional Election Study Data: We analyzed responses from the 2012 CCES, a survey of U.S. citizens aged 18 and over, to obtain state-level estimates of wait times reported by voters for in-person voting on Election Day 2012.[Footnote 95] Specifically, CCES respondents were asked to estimate wait times within specified response categories for the 2012 general election. To estimate voter wait times on Election Day 2012, we replicated an approach used by another researcher and estimated average wait times by first recoding the response categories to the midpoint of the category--for example, the "none at all" response was coded as 0 minutes, and the "1-10" response category was coded as 5 minutes. [Footnote 96] Respondents who waited more than an hour were asked to provide wait times in minutes. To assess the reliability of these data, we reviewed documentation related to the 2012 CCES and interviewed researchers knowledgeable about the survey. We determined that the CCES data used in this report were sufficiently reliable for our purposes. Interviews with State and Local Jurisdiction Election Officials: We interviewed state election officials from 47 states and the District of Columbia to obtain such information as the availability of data on voter wait times in their states for Election Day 2012 and their views on policies and procedures that may have affected voter wait times.[Footnote 97] Because of differences in election administration across states, these officials were located in various state offices, including state secretary of state or commonwealth offices, boards of elections, and lieutenant governors' offices. We corroborated the information we gathered through these interviews by reviewing any documentation that these states provided, such as guidance on planning elections and voter wait time reports. We also interviewed local election officials, on-site or by phone, from 5 local jurisdictions--Detroit, Michigan; Hartford, Connecticut; Lee County, Florida; Los Angeles County, California; and Prince William County, Virginia--to perform a more detailed examination of their experiences on Election Day 2012, including how, if at all, they measured wait times, their views on the factors that affected wait times in their respective jurisdictions on Election Day 2012, and their perspectives on the specific impacts of these factors, among other things. We selected these jurisdictions (1) to reflect variation in geographic location and demographic characteristics, and (2) based on our survey results, CCES results, and our review of the wait time literature, to include a range of voter wait times and election administration policies and practices. For example, in our survey, 4 of the 5 selected jurisdictions reported having varying extents of long wait times and 1 reported not having long wait times. In each jurisdiction, we interviewed the chief election official; other officials from the elections office; and, if available, individuals who had served as poll workers at polling locations in the jurisdiction on Election Day 2012. While these 5 jurisdictions are not representative of all election jurisdictions nationwide and their responses cannot be generalized to other local election jurisdictions, officials in these locations provided a range of perspectives on voter wait times and information on how factors affected wait times in practice and allowed us to compare Election Day 2012 experiences across jurisdictions. We corroborated the information we gathered through these interviews by reviewing postelection reports, relevant state statutes, and documentation that these jurisdictions provided to us, such as data relating to voter wait times and poll worker training materials. We interviewed officials from these jurisdictions between May and July 2014. Interviews with Election Administration Researchers: We also interviewed officials from the Election Assistance Commission and 14 researchers and representatives from research organizations in the field of election administration to discuss their research and perspectives on wait time measurement and voter wait times.[Footnote 98] We selected these researchers and representatives based on our review of voter wait time literature, their expertise and work in this area, and recommendations from these and other researchers. The information that we obtained cannot be generalized to other researchers; however, these interviews provided a range of views on such areas as practices for measuring wait times, the frequency of long voter wait times, and factors affecting wait times. We conducted this performance audit from July 2013 to September 2014 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. [End of section] Appendix II: Aggregated Results from Nationwide Survey of Local Election Jurisdictions: The questions we asked in our survey of local election jurisdictions are shown below. Our survey was composed of closed-and open-ended questions. In this appendix, we include all survey questions and aggregate results of responses to the closed-ended questions; we do not provide information on responses provided to the open-ended questions. For a more detailed discussion of our survey methodology, see appendix I. Definitions: Throughout this questionnaire, we use certain terms. For example: The term "your jurisdiction" means your local election jurisdiction. Also, when we refer to "voter wait time", we mean the time from when a voter entered the first line to when they began filling out a ballot. We recognize that the time spent filling out or submitting a ballot may affect the wait time of later voters in line, but we would like you to consider the voter wait time to be only the time a voter waits prior to filling out a ballot. We have also included additional questions in this questionnaire about the time it took to turn in a ballot. 1. What is the name, title, and telephone number of the primary person completing this questionnaire so that we may contact someone if we need to clarify any responses? Name: Title: Telephone number: Email: Part I: Local Jurisdiction Characteristics: 2. Approximately how many polling places and precincts were there in your jurisdiction on the November 2012 General Election Day? [open ended] Number of polling places: Number of precincts: 3. On the November 2012 General Election Day, what was the total number of all registered voters in your jurisdiction? [open ended] Number of all registered voters: 4. Did your jurisdiction collect, receive, or have available information that would allow you to calculate or estimate voter wait times that occurred at individual polling places on the November 2012 General Election Day? (Please check one response): Response: Yes (Skip to question 6); Estimated percentage: 16.0%; 95 percent confidence interval-lower bound (percentage): 12.0%; 95 percent confidence interval-upper bound (percentage): 20.7%. Response: No (Go to question 5); Estimated percentage: 78.3%; 95 percent confidence interval-lower bound (percentage): 73.5%; 95 percent confidence interval-upper bound (percentage): 83.1%. Response: Don't know (Skip to question 6); Estimated percentage: 5.7%; 95 percent confidence interval-lower bound (percentage): 3.3%; 95 percent confidence interval-upper bound (percentage): 9.1%. 5. IF No: Which of the following, if any, were reasons your jurisdiction did NOT collect, receive, or have available information on voter wait time? (Check all that apply): Response: 5a. Too expensive to collect; Estimated percentage: 2.8%; 95 percent confidence interval-lower bound (percentage): 1.1%; 95 percent confidence interval-upper bound (percentage): 5.8%. Response: 5b. Unsure how to collect such data; Estimated percentage: 9.1%; 95 percent confidence interval-lower bound (percentage): 5.7%; 95 percent confidence interval-upper bound (percentage): 13.7%. Response: 5c. Data required to calculate voter wait times were not available; Estimated percentage: 10.4%; 95 percent confidence interval-lower bound (percentage): 6.7%; 95 percent confidence interval-upper bound (percentage): 15.2%. Response: 5d. No requirements existed to collect such data; Estimated percentage: 46.8%; 95 percent confidence interval-lower bound (percentage): 40.2%; 95 percent confidence interval-upper bound (percentage): 53.3%. Response: 5e. Voter wait time has not been an issue; Estimated percentage: 78.8%; 95 percent confidence interval-lower bound (percentage): 73.4%; 95 percent confidence interval-upper bound (percentage): 84.1%. Response: 5f. Voter wait times were not a specified goal of our Election Day activities; Estimated percentage: 14.1%; 95 percent confidence interval-lower bound (percentage): 10.0%; 95 percent confidence interval-upper bound (percentage): 19.1%. Response: 5g. Other (please specify below); Estimated percentage: 2.1%; 95 percent confidence interval-lower bound (percentage): 0.7%; 95 percent confidence interval-upper bound (percentage): 4.6%. For what other reason(s) did your jurisdiction not collect, receive, or have available information on voter wait times? [open ended] 6. Did your jurisdiction collect, receive, or have available any of the following information for the November 2012 General Election Day? (Check one response on each row): 6a. The time individuals checked into a polling place recorded by an electronic poll book: Response: Yes; Estimated percentage: 17.1; 95 percent confidence interval-lower bound (percentage): 12.8; 95 percent confidence interval-upper bound (percentage): 22.1. Response: No; Estimated percentage: 63.1; 95 percent confidence interval-lower bound (percentage): 57.4; 95 percent confidence interval-upper bound (percentage): 68.8. Response: Not applicable; Estimated percentage: 17.0; 95 percent confidence interval-lower bound (percentage): 12.9; 95 percent confidence interval-upper bound (percentage): 21.9. Response: No response; Estimated percentage: 2.8; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 5.5. 6b. The time individuals checked into a polling place, recorded by a method other than an electronic poll book: Response: Yes; Estimated percentage: 5.1; 95 percent confidence interval-lower bound (percentage): 2.9; 95 percent confidence interval-upper bound (percentage): 8.3. Response: No; Estimated percentage: 79.5; 95 percent confidence interval-lower bound (percentage): 74.8; 95 percent confidence interval-upper bound (percentage): 84.1. Response: Not applicable; Estimated percentage: 12.7; 95 percent confidence interval-lower bound (percentage): 9.1; 95 percent confidence interval-upper bound (percentage): 17.2. Response: No response; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 5.3. Response: 6c. The number of votes cast at a polling place during specific time periods: Response: Yes; Estimated percentage: 30.7; 95 percent confidence interval-lower bound (percentage): 25.3; 95 percent confidence interval-upper bound (percentage): 36.0. Response: No; Estimated percentage: 57.7; 95 percent confidence interval-lower bound (percentage): 51.9; 95 percent confidence interval-upper bound (percentage): 63.5. Response: Not applicable; Estimated percentage: 8.6; 95 percent confidence interval-lower bound (percentage): 5.6; 95 percent confidence interval-upper bound (percentage): 12.5. Response: No response; Estimated percentage: 3.1; 95 percent confidence interval-lower bound (percentage): 1.3; 95 percent confidence interval-upper bound (percentage): 6.0. Response: 6d. The time votes were cast based on the voting machine time stamp: Response: Yes; Estimated percentage: 14.5; 95 percent confidence interval-lower bound (percentage): 10.6; 95 percent confidence interval-upper bound (percentage): 19.0. Response: No; Estimated percentage: 64.3; 95 percent confidence interval-lower bound (percentage): 58.7; 95 percent confidence interval-upper bound (percentage): 69.8. Response: Not applicable; Estimated percentage: 15.7; 95 percent confidence interval-lower bound (percentage): 11.6; 95 percent confidence interval-upper bound (percentage): 20.5. Response: No response; Estimated percentage: 5.6; 95 percent confidence interval-lower bound (percentage): 3.3; 95 percent confidence interval-upper bound (percentage): 9.0. Response: 6e. Voter complaints about wait times at polling places; Response: Yes; Estimated percentage: 16.3; 95 percent confidence interval-lower bound (percentage): 12.4; 95 percent confidence interval-upper bound (percentage): 20.9. Response: No; Estimated percentage: 66.2; 95 percent confidence interval-lower bound (percentage): 60.8; 95 percent confidence interval-upper bound (percentage): 71.7. Response: Not applicable; Estimated percentage: 13.5; 95 percent confidence interval-lower bound (percentage): 9.7; 95 percent confidence interval-upper bound (percentage): 18.1. Response: No response; Estimated percentage: 3.9; 95 percent confidence interval-lower bound (percentage): 1.9; 95 percent confidence interval-upper bound (percentage): 7.1. Response: 6f. The length of time polling places remained open after designated closing times: Response: Yes; Estimated percentage: 17.9; 95 percent confidence interval-lower bound (percentage): 13.6; 95 percent confidence interval-upper bound (percentage): 22.9. Response: No; Estimated percentage: 52.2; 95 percent confidence interval-lower bound (percentage): 46.3; 95 percent confidence interval-upper bound (percentage): 58.2. Response: Not applicable; Estimated percentage: 26.7; 95 percent confidence interval-lower bound (percentage): 21.4; 95 percent confidence interval-upper bound (percentage): 31.9. Response: No response; Estimated percentage: 3.2; 95 percent confidence interval-lower bound (percentage): 1.5; 95 percent confidence interval-upper bound (percentage): 6.0. Response: 6g. Observations of voter wait times at polling places by election officials: Response: Yes; Estimated percentage: 36.1; 95 percent confidence interval-lower bound (percentage): 30.5; 95 percent confidence interval-upper bound (percentage): 41.8. Response: No; Estimated percentage: 53.0; 95 percent confidence interval-lower bound (percentage): 47.2; 95 percent confidence interval-upper bound (percentage): 58.9. Response: Not applicable; Estimated percentage: 7.3; 95 percent confidence interval-lower bound (percentage): 4.6; 95 percent confidence interval-upper bound (percentage): 10.9. Response: No response; Estimated percentage: 3.5; 95 percent confidence interval-lower bound (percentage): 1.7; 95 percent confidence interval-upper bound (percentage): 6.4. Response: 6h. Other information related to voter wait time (please specify below): Response: Yes; Estimated percentage: 4.6; 95 percent confidence interval-lower bound (percentage): 2.5; 95 percent confidence interval-upper bound (percentage): 7.7. Response: No; Estimated percentage: 63.2; 95 percent confidence interval-lower bound (percentage): 57.4; 95 percent confidence interval-upper bound (percentage): 69.0. Response: Not applicable; Estimated percentage: 21.3; 95 percent confidence interval-lower bound (percentage): 16.4; 95 percent confidence interval-upper bound (percentage): 26.3. Response: No response; Estimated percentage: 0.9; 95 percent confidence interval-lower bound (percentage): 7.4; 95 percent confidence interval-upper bound (percentage): 15.2. [End of table] What other information did your jurisdiction collect, receive, or have available on voter wait time? [open ended] 7. Did your jurisdiction make a formal calculation of voter wait times that occurred on the November 2012 General Election Day? (Please check one response): Response: Yes (Go to question 8); Estimated percentage: 4.2; 95 percent confidence interval-lower bound (percentage): 2.2; 95 percent confidence interval-upper bound (percentage): 7.2. Response: No (Skip to question 9); Estimated percentage: 94.0; 95 percent confidence interval-lower bound (percentage): 90.5; 95 percent confidence interval-upper bound (percentage): 96.5. Response: Don't know (Skip to question 9); Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 4.3. [End of table] 8. How did you calculate voter wait times? Please include the types of data you collected that you used to calculate voter wait times as well as the method you used to analyze the data. [open-ended] 9. What type of polling place level data, if any, did your jurisdiction use when allocating resources for the November 2012 General Election Day? (Check one response on each row): 9a. Registered voters: Response: Specific data for each polling place; Estimated percentage: 78.3; 95 percent confidence interval-lower bound (percentage): 73.5; 95 percent confidence interval-upper bound (percentage): 83.1. Response: Estimate for each polling place; Estimated percentage: 10.4; 95 percent confidence interval-lower bound (percentage): 7.1; 95 percent confidence interval-upper bound (percentage): 14.7. Response: None used; Estimated percentage: 7.4; 95 percent confidence interval-lower bound (percentage): 4.6; 95 percent confidence interval-upper bound (percentage): 11.1. Response: No response; Estimated percentage: 3.9; 95 percent confidence interval-lower bound (percentage): 1.9; 95 percent confidence interval-upper bound (percentage): 6.9. 9b. Active voters; Response: Specific data for each polling place; Estimated percentage: 69.3; 95 percent confidence interval-lower bound (percentage): 63.8; 95 percent confidence interval-upper bound (percentage): 74.7. Response: Estimate for each polling place; Estimated percentage: 13.9; 95 percent confidence interval-lower bound (percentage): 10.0; 95 percent confidence interval-upper bound (percentage): 18.6. Response: None used; Estimated percentage: 12.0; 95 percent confidence interval-lower bound (percentage): 8.4; 95 percent confidence interval-upper bound (percentage): 16.5. Response: No response; Estimated percentage: 4.8; 95 percent confidence interval-lower bound (percentage): 2.5; 95 percent confidence interval-upper bound (percentage): 8.0. 9c. Voter turnout from previous presidential general election(s); Response: Specific data for each polling place; Estimated percentage: 60.8; 95 percent confidence interval-lower bound (percentage): 55.1; 95 percent confidence interval-upper bound (percentage): 66.5. Response: Estimate for each polling place; Estimated percentage: 17.8; 95 percent confidence interval-lower bound (percentage): 13.5; 95 percent confidence interval-upper bound (percentage): 22.9. Response: None used; Estimated percentage: 16.5; 95 percent confidence interval-lower bound (percentage): 12.3; 95 percent confidence interval-upper bound (percentage): 21.5. Response: No response; Estimated percentage: 4.9; 95 percent confidence interval-lower bound (percentage): 2.7; 95 percent confidence interval-upper bound (percentage): 8.2. 9d. Voter turnout by hour from previous presidential general election(s); Response: Specific data for each polling place; Estimated percentage: 7.1; 95 percent confidence interval-lower bound (percentage): 4.4; 95 percent confidence interval-upper bound (percentage): 10.8. Response: Estimate for each polling place; Estimated percentage: 8.1; 95 percent confidence interval-lower bound (percentage): 5.3; 95 percent confidence interval-upper bound (percentage): 11.8. Response: None used; Estimated percentage: 76.4; 95 percent confidence interval-lower bound (percentage): 71.4; 95 percent confidence interval-upper bound (percentage): 81.4. Response: No response; Estimated percentage: 8.4; 95 percent confidence interval-lower bound (percentage): 5.4; 95 percent confidence interval-upper bound (percentage): 12.3. Response: 9e. Number of voters expected to vote or who actually voted outside of Election Day (e.g., early, absentee, or mail-in voting); Response: Specific data for each polling place; Estimated percentage: 40.1; 95 percent confidence interval-lower bound (percentage): 34.3; 95 percent confidence interval-upper bound (percentage): 45.9. Response: Estimate for each polling place; Estimated percentage: 25.9; 95 percent confidence interval-lower bound (percentage): 20.7; 95 percent confidence interval-upper bound (percentage): 31.1. Response: None used; Estimated percentage: 27.7; 95 percent confidence interval-lower bound (percentage): 22.3; 95 percent confidence interval-upper bound (percentage): 33.0. Response: No response; Estimated percentage: 6.3; 95 percent confidence interval-lower bound (percentage): 3.7; 95 percent confidence interval-upper bound (percentage): 10.0. 9f. Average time needed to check in voters; Response: Specific data for each polling place; Estimated percentage: 9.9; 95 percent confidence interval-lower bound (percentage): 6.7; 95 percent confidence interval-upper bound (percentage): 14.0. Response: Estimate for each polling place; Estimated percentage: 20.7; 95 percent confidence interval-lower bound (percentage): 16.0; 95 percent confidence interval-upper bound (percentage): 25.4. Response: None used; Estimated percentage: 61.2; 95 percent confidence interval-lower bound (percentage): 55.4; 95 percent confidence interval-upper bound (percentage): 67.0. Response: No response; Estimated percentage: 8.2; 95 percent confidence interval-lower bound (percentage): 5.2; 95 percent confidence interval-upper bound (percentage): 12.2. 9g. Average time needed to complete ballots; Response: Specific data for each polling place; Estimated percentage: 11.6; 95 percent confidence interval-lower bound (percentage): 8.0; 95 percent confidence interval-upper bound (percentage): 16.0. Response: Estimate for each polling place; Estimated percentage: 23.6; 95 percent confidence interval-lower bound (percentage): 18.6; 95 percent confidence interval-upper bound (percentage): 28.6. Response: None used; Estimated percentage: 56.3; 95 percent confidence interval-lower bound (percentage): 50.4; 95 percent confidence interval-upper bound (percentage): 62.1. Response: No response; Estimated percentage: 8.6; 95 percent confidence interval-lower bound (percentage): 5.5; 95 percent confidence interval-upper bound (percentage): 12.6. 9h. Demographics (including presence of non-English-speaking voters, elderly population, etc.); Response: Specific data for each polling place; Estimated percentage: 7.9; 95 percent confidence interval-lower bound (percentage): 5.0; 95 percent confidence interval-upper bound (percentage): 11.6. Response: Estimate for each polling place; Estimated percentage: 14.8; 95 percent confidence interval-lower bound (percentage): 10.9; 95 percent confidence interval-upper bound (percentage): 19.5. Response: None used; Estimated percentage: 66.9; 95 percent confidence interval-lower bound (percentage): 61.4; 95 percent confidence interval-upper bound (percentage): 72.5. Response: No response; Estimated percentage: 10.4; 95 percent confidence interval-lower bound (percentage): 7.0; 95 percent confidence interval-upper bound (percentage): 14.7. 9i. Other (please specify below); Response: Specific data for each polling place; Estimated percentage: 2.9; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 6.3. Response: Estimate for each polling place; Estimated percentage: 1.9; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 5.0. Response: None used; Estimated percentage: 52.6; 95 percent confidence interval-lower bound (percentage): 45.2; 95 percent confidence interval-upper bound (percentage): 60.0. Response: No response; Estimated percentage: 42.6; 95 percent confidence interval-lower bound (percentage): 35.2; 95 percent confidence interval-upper bound (percentage): 50.0. [End of table] What other polling place level data did your jurisdiction use? [open ended] Part II: Voter Wait Time on General Election Day November 2012: 10. On average for all of the polling places in your jurisdiction for the November 2012 General Election Day, how long did it typically take for a voter to wait to begin filling out a ballot during the following times of day? Please consider the time from when a voter entered the first line to when they began filling out a ballot. (Please check one response on each column): Response: 5 minutes or less; First hour after polls opened; Estimated percentage: 32.1; 95 percent confidence interval-lower bound (percentage): 26.7; 95 percent confidence interval-upper bound (percentage): 37.6. Around lunchtime; Estimated percentage: 26.8; 95 percent confidence interval-lower bound (percentage): 21.6; 95 percent confidence interval-upper bound (percentage): 32.0. Last hour before polls closed; Estimated percentage: 24.8; 95 percent confidence interval-lower bound (percentage): 19.7; 95 percent confidence interval-upper bound (percentage): 29.8. Response: 6 to 10 minutes; First hour after polls opened; Estimated percentage: 18.4; 95 percent confidence interval-lower bound (percentage): 13.9; 95 percent confidence interval-upper bound (percentage): 22.9. Around lunchtime; Estimated percentage: 22.5; 95 percent confidence interval-lower bound (percentage): 17.7; 95 percent confidence interval-upper bound (percentage): 27.3. Last hour before polls closed; Estimated percentage: 20.5; 95 percent confidence interval-lower bound (percentage): 15.7; 95 percent confidence interval-upper bound (percentage): 25.3. Response: 11 to 20 minutes; First hour after polls opened; Estimated percentage: 8.5; 95 percent confidence interval-lower bound (percentage): 5.6; 95 percent confidence interval-upper bound (percentage): 12.3. Around lunchtime; Estimated percentage: 12.1; 95 percent confidence interval-lower bound (percentage): 8.5; 95 percent confidence interval-upper bound (percentage): 16.5. Last hour before polls closed; Estimated percentage: 13.7; 95 percent confidence interval-lower bound (percentage): 9.9; 95 percent confidence interval-upper bound (percentage): 18.3. Response: 21 to 30 minutes; First hour after polls opened; Estimated percentage: 5.2; 95 percent confidence interval-lower bound (percentage): 3.0; 95 percent confidence interval-upper bound (percentage): 8.5. Response: Around lunchtime; Estimated percentage: 2.3; 95 percent confidence interval-lower bound (percentage): 0.9; 95 percent confidence interval-upper bound (percentage): 4.7. Response: Last hour before polls closed; Estimated percentage: 4.5; 95 percent confidence interval-lower bound (percentage): 2.3; 95 percent confidence interval-upper bound (percentage): 7.6. Response: 31 to 60 minutes; First hour after polls opened; Estimated percentage: 1.9; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 4.2. Around lunchtime; Estimated percentage: 2.3; 95 percent confidence interval-lower bound (percentage): 0.9; 95 percent confidence interval-upper bound (percentage): 5.0. Last hour before polls closed; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 5.2. Response: 61 to 120 minutes; First hour after polls opened; Estimated percentage: 0.9; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 3.0. Around lunchtime; Estimated percentage: 0.1; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.7. Last hour before polls closed; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 3.2. Response: More than 120 minutes; First hour after polls opened; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.7. Around lunchtime; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. Last hour before polls closed; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.7. Response: Don't know; First hour after polls opened; Estimated percentage: 30.3; 95 percent confidence interval-lower bound (percentage): 25.0; 95 percent confidence interval-upper bound (percentage): 35.7. Around lunchtime; Estimated percentage: 31.2; 95 percent confidence interval-lower bound (percentage): 25.8; 95 percent confidence interval-upper bound (percentage): 36.7. Last hour before polls closed; Estimated percentage: 30.2; 95 percent confidence interval-lower bound (percentage): 24.8; 95 percent confidence interval-upper bound (percentage): 35.5. Response: No response; First hour after polls opened; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 5.2. Around lunchtime; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 5.2. Last hour before polls closed; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 5.2. [End of table] 11. Would you say that the voter wait times you described in the previous question were greater than, about the same as, or less than the typical voter wait times for the November 2008 General Election Day? (Please check one response): Response: Greater than the typical November 2008 General Election Day voter wait times; Estimated percentage: 5.6; 95 percent confidence interval-lower bound (percentage): 3.3; 95 percent confidence interval-upper bound (percentage): 8.9. Response: Less than the typical November 2008 General Election Day voter wait times; Estimated percentage: 19.0; 95 percent confidence interval-lower bound (percentage): 14.4; 95 percent confidence interval-upper bound (percentage): 23.6. Response: About the same as the typical November 2008 General Election Day voter wait times; Estimated percentage: 44.3; 95 percent confidence interval-lower bound (percentage): 38.4; 95 percent confidence interval-upper bound (percentage): 50.1. Response: Don't know/unsure; Estimated percentage: 31.1; 95 percent confidence interval-lower bound (percentage): 25.6; 95 percent confidence interval-upper bound (percentage): 36.6. [End of table] 12. In your opinion, what would you consider to be a voter wait time on Election Day that is too long? (Please check one response): Response: 10 minutes or less; Estimated: percentage: 7.2; 95 percent confidence interval-lower bound (percentage): 4.4; 95 percent confidence interval-upper bound (percentage): 10.9. Response: More than 10 minutes; Estimated: percentage: 23.9; 95 percent confidence interval-lower bound (percentage): 18.9; 95 percent confidence interval-upper bound (percentage): 28.9. Response: More than 20 minutes; Estimated: percentage: 29.7; 95 percent confidence interval-lower bound (percentage): 24.3; 95 percent confidence interval-upper bound (percentage): 35.1. Response: More than 30 minutes; Estimated: percentage: 21.5; 95 percent confidence interval-lower bound (percentage): 16.7; 95 percent confidence interval-upper bound (percentage): 26.2. Response: More than 60 minutes; Estimated: percentage: 10.7; 95 percent confidence interval-lower bound (percentage): 7.3; 95 percent confidence interval-upper bound (percentage): 14.9. Response: More than 120 minutes; Estimated: percentage: 0.7; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.4. Response: Don't know; Estimated: percentage: 6.4; 95 percent confidence interval-lower bound (percentage): 3.8; 95 percent confidence interval-upper bound (percentage): 9.9. [End of table] 13. How many polling places in your jurisdiction had voter wait times that were too long on the November 2012 General Election Day? (Please check one response): Response: No polling places (Skip to question 15); Estimated: percentage: 78.0; 95 percent confidence interval-lower bound (percentage): 73.4; 95 percent confidence interval-upper bound (percentage): 82.6. Response: Only a few polling places (Go to question 14); Estimated: percentage: 19.0; 95 percent confidence interval-lower bound (percentage): 14.6; 95 percent confidence interval-upper bound (percentage): 23.5. Response: Less than half, but more than a few polling places (Go to question 14); Estimated: percentage: 2.3; 95 percent confidence interval-lower bound (percentage): 0.9; 95 percent confidence interval-upper bound (percentage): 4.5. Response: About half of all polling places (Go to question 14); Estimated: percentage: 0.1; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.7. Response: More than half of all polling places (Go to question 14); Estimated: percentage: 0.6; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.1. [End of table] 14. In thinking about the November 2012 General Election Day, how much of a factor, if any, do you believe each of the following was to long voter wait times at polling places in your jurisdiction? Response: 14a. No or limited opportunities for voting outside of Election Day (e.g., early or mail-in voting); Major factor; Estimated percentage: 3.3; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 9.8. Minor factor; Estimated percentage: 21.2; 95 percent confidence interval-lower bound (percentage): 12.0; 95 percent confidence interval-upper bound (percentage): 33.2. Not a factor; Estimated percentage: 65.5; 95 percent confidence interval-lower bound (percentage): 54.2; 95 percent confidence interval-upper bound (percentage): 76.9. Not applicable; Estimated percentage: 8.2; 95 percent confidence interval-lower bound (percentage): 2.9; 95 percent confidence interval-upper bound (percentage): 17.6. Response: No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14b. Not enough locations for in-person voting prior to Election Day; Major factor; Estimated percentage: 0.5; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. Minor factor; Estimated percentage: 19.3; 95 percent confidence interval-lower bound (percentage): 10.4; 95 percent confidence interval-upper bound (percentage): 31.3. Not a factor; Estimated percentage: 62.0; 95 percent confidence interval-lower bound (percentage): 50.2; 95 percent confidence interval-upper bound (percentage): 73.9. Not applicable; Estimated percentage: 16.4; 95 percent confidence interval-lower bound (percentage): 8.5; 95 percent confidence interval-upper bound (percentage): 27.5. No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14c. Not enough days or hours for in-person voting prior to Election Day; Major factor; Estimated percentage: 0.6; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.2. Minor factor; Estimated percentage: 11.6; 95 percent confidence interval-lower bound (percentage): 4.8; 95 percent confidence interval-upper bound (percentage): 22.4. Not a factor; Estimated percentage: 73.3; 95 percent confidence interval-lower bound (percentage): 60.8; 95 percent confidence interval-upper bound (percentage): 83.5. Not applicable; Estimated percentage: 12.8; 95 percent confidence interval-lower bound (percentage): 5.8; 95 percent confidence interval-upper bound (percentage): 23.5. No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14d. Not enough Election Day polling places; Major factor; Estimated percentage: 0.6; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. Response: Minor factor; Estimated percentage: 13.0; 95 percent confidence interval-lower bound (percentage): 5.9; 95 percent confidence interval-upper bound (percentage): 23.7. Not a factor; Estimated percentage: 82.3; 95 percent confidence interval-lower bound (percentage): 70.9; 95 percent confidence interval-upper bound (percentage): 90.7. Not applicable; Estimated percentage: 2.2; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 11.0. No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14e. Design/layout of polling places; Major factor; Estimated percentage: 4.6; 95 percent confidence interval-lower bound (percentage): 0.9; 95 percent confidence interval-upper bound (percentage): 13.1. Minor factor; Estimated percentage: 28.1; 95 percent confidence interval-lower bound (percentage): 18.0; 95 percent confidence interval-upper bound (percentage): 40.1. Not a factor; Estimated percentage: 65.3; 95 percent confidence interval-lower bound (percentage): 54.1; 95 percent confidence interval-upper bound (percentage): 76.5. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 2.0; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 8.9. Response: 14f. Not enough voting machines; Major factor; Estimated percentage: 8.9; 95 percent confidence interval-lower bound (percentage): 3.0; 95 percent confidence interval-upper bound (percentage): 19.6. Minor factor; Estimated percentage: 18.3; 95 percent confidence interval-lower bound (percentage): 10.2; 95 percent confidence interval-upper bound (percentage): 29.0. Not a factor; Estimated percentage: 72.8; 95 percent confidence interval-lower bound (percentage): 60.6; 95 percent confidence interval-upper bound (percentage): 83.0. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14g. Voting machine failures; Major factor; Estimated percentage: 0.3; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. Minor factor; Estimated percentage: 15.9; 95 percent confidence interval-lower bound (percentage): 8.2; 95 percent confidence interval-upper bound (percentage): 26.8. Not a factor; Estimated percentage: 82.1; 95 percent confidence interval-lower bound (percentage): 70.9; 95 percent confidence interval-upper bound (percentage): 90.3. Not applicable; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 8.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14h. Type of voting method or machine; Major factor; Estimated percentage: 5.2; 95 percent confidence interval-lower bound (percentage): 1.3; 95 percent confidence interval-upper bound (percentage): 13.4. Minor factor; Estimated percentage: 5.6; 95 percent confidence interval-lower bound (percentage): 1.8; 95 percent confidence interval-upper bound (percentage): 12.6. Not a factor; Estimated percentage: 89.3; 95 percent confidence interval-lower bound (percentage): 80.0; 95 percent confidence interval-upper bound (percentage): 95.2. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14i. Not enough poll workers at polling places; Major factor; Estimated percentage: 2.8; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 8.5. Minor factor; Estimated percentage: 22.9; 95 percent confidence interval-lower bound (percentage): 13.7; 95 percent confidence interval-upper bound (percentage): 34.7. Not a factor; Estimated percentage: 74.3; 95 percent confidence interval-lower bound (percentage): 62.4; 95 percent confidence interval-upper bound (percentage): 84.0. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14j. Training of poll workers; Major factor; Estimated percentage: 2.6; 95 percent confidence interval-lower bound (percentage): 0.4; 95 percent confidence interval-upper bound (percentage): 8.6. Minor factor; Estimated percentage: 33.6; 95 percent confidence interval-lower bound (percentage): 22.5; 95 percent confidence interval-upper bound (percentage): 46.2. Not a factor; Estimated percentage: 63.9; 95 percent confidence interval-lower bound (percentage): 52.3; 95 percent confidence interval-upper bound (percentage): 75.4. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14k. Higher than expected voter turnout; Major factor; Estimated percentage: 23.2; 95 percent confidence interval-lower bound (percentage): 13.3; 95 percent confidence interval-upper bound (percentage): 35.7. Minor factor; Estimated percentage: 21.9; 95 percent confidence interval-lower bound (percentage): 13.2; 95 percent confidence interval-upper bound (percentage): 32.7. Not a factor; Estimated percentage: 55.0; 95 percent confidence interval-lower bound (percentage): 42.9; 95 percent confidence interval-upper bound (percentage): 67.0. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14l. Redistricting; Major factor; Estimated percentage: 13.2; 95 percent confidence interval-lower bound (percentage): 6.2; 95 percent confidence interval-upper bound (percentage): 23.7. Minor factor; Estimated percentage: 21.7; 95 percent confidence interval-lower bound (percentage): 12.6; 95 percent confidence interval-upper bound (percentage): 33.4. Not a factor; Estimated percentage: 62.7; 95 percent confidence interval-lower bound (percentage): 51.1; 95 percent confidence interval-upper bound (percentage): 74.2. Not applicable; Estimated percentage: 2.4; 95 percent confidence interval-lower bound (percentage): Response: 0.1; 95 percent confidence interval-upper bound (percentage): Response: 10.8. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14m. Consolidation/changes in polling places; Major factor; Estimated percentage: 7.7; 95 percent confidence interval-lower bound (percentage): 2.6; 95 percent confidence interval-upper bound (percentage): 16.9. Minor factor; Estimated percentage: 20.5; 95 percent confidence interval-lower bound (percentage): 11.7; 95 percent confidence interval-upper bound (percentage): 32.0. Not a factor; Estimated percentage: 68.1; 95 percent confidence interval-lower bound (percentage): 55.6; 95 percent confidence interval-upper bound (percentage): 79.0. Not applicable; Estimated percentage: 3.7; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 11.9. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14n. Use of paper poll books; Major factor; Estimated percentage: 10.4; 95 percent confidence interval-lower bound (percentage): 4.4; 95 percent confidence interval-upper bound (percentage): 19.9. Minor factor; Estimated percentage: 24.2; 95 percent confidence interval-lower bound (percentage): 14.3; 95 percent confidence interval-upper bound (percentage): 36.6. Not a factor; Estimated percentage: 47.7; 95 percent confidence interval-lower bound (percentage): 35.7; 95 percent confidence interval-upper bound (percentage): 59.6. Not applicable; Estimated percentage: 17.7; 95 percent confidence interval-lower bound (percentage): 9.4; 95 percent confidence interval-upper bound (percentage): 29.2. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14o. Use of electronic poll books; Major factor; Estimated percentage: 0.5; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. Minor factor; Estimated percentage: 14.6; 95 percent confidence interval-lower bound (percentage): 6.9; 95 percent confidence interval-upper bound (percentage): 25.8. Not a factor; Estimated percentage: 43.7; 95 percent confidence interval-lower bound (percentage): 31.5; 95 percent confidence interval-upper bound (percentage): 55.9. Not applicable; Estimated percentage: 41.0; 95 percent confidence interval-lower bound (percentage): 29.2; 95 percent confidence interval-upper bound (percentage): 52.8. No response; Estimated percentage: 0.2; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. Response: 14p. Incorrect/inaccurate information on voter registration rolls; Major factor; Estimated percentage: 2.7; 95 percent confidence interval-lower bound (percentage): 0.4; 95 percent confidence interval-upper bound (percentage): 8.7. Minor factor; Estimated percentage: 21.0; 95 percent confidence interval-lower bound (percentage): 11.8; 95 percent confidence interval-upper bound (percentage): 33.0. Not a factor; Estimated percentage: 70.2; 95 percent confidence interval-lower bound (percentage): 57.7; 95 percent confidence interval-upper bound (percentage): 80.8. Not applicable; Estimated percentage: 6.1; 95 percent confidence interval-lower bound (percentage): 1.6; 95 percent confidence interval-upper bound (percentage): 15.1. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14q. Inadequate number of personnel to handle calls from poll workers when problems come up at polling places; Major factor; Estimated percentage: 7.5; 95 percent confidence interval-lower bound (percentage): 2.2; 95 percent confidence interval-upper bound (percentage): 17.5. Minor factor; Estimated percentage: 30.6; 95 percent confidence interval-lower bound (percentage): 19.8; 95 percent confidence interval-upper bound (percentage): 43.2. Not a factor; Estimated percentage: 61.7; 95 percent confidence interval-lower bound (percentage): 50.0; 95 percent confidence interval-upper bound (percentage): 73.4. Not applicable; Estimated percentage: 0.2; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14r. Election Day voter registration; Major factor; Estimated percentage: 3.2; 95 percent confidence interval-lower bound (percentage): 0.4; 95 percent confidence interval-upper bound (percentage): 10.8. Minor factor; Estimated percentage: 10.7; 95 percent confidence interval-lower bound (percentage): 4.3; 95 percent confidence interval-upper bound (percentage): 21.3. Not a factor; Estimated percentage: 37.0; 95 percent confidence interval-lower bound (percentage): 25.5; 95 percent confidence interval-upper bound (percentage): 48.5. Not applicable; Estimated percentage: 49.1; 95 percent confidence interval-lower bound (percentage): 37.1; 95 percent confidence interval-upper bound (percentage): 61.0. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14s. Large number of first-time voters; Major factor; Estimated percentage: 6.0; 95 percent confidence interval-lower bound (percentage): 1.7; 95 percent confidence interval-upper bound (percentage): 14.9. Minor factor; Estimated percentage: 43.0; 95 percent confidence interval-lower bound (percentage): 31.0; 95 percent confidence interval-upper bound (percentage): 55.1. Not a factor; Estimated percentage: 49.1; 95 percent confidence interval-lower bound (percentage): 37.1; 95 percent confidence interval-upper bound (percentage): 61.1. Not applicable; Estimated percentage: Response: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14t. Large number of inactive voters; Major factor; Estimated percentage: 12.1; 95 percent confidence interval-lower bound (percentage): 5.2; 95 percent confidence interval-upper bound (percentage): 22.8. Minor factor; Estimated percentage: 23.1; 95 percent confidence interval-lower bound (percentage): 13.5; 95 percent confidence interval-upper bound (percentage): 35.2. Not a factor; Estimated percentage: 63.0; 95 percent confidence interval-lower bound (percentage): 51.3; 95 percent confidence interval-upper bound (percentage): 74.8. Not applicable; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 9.1. Response: 14u. Long ballot; Major factor; Estimated percentage: 31.4; 95 percent confidence interval-lower bound (percentage): 20.6; 95 percent confidence interval-upper bound (percentage): 44.0. Minor factor; Estimated percentage: 39.8; 95 percent confidence interval-lower bound (percentage): 27.8; 95 percent confidence interval-upper bound (percentage): 51.9. Not a factor; Estimated percentage: 26.4; 95 percent confidence interval-lower bound (percentage): 16.4; 95 percent confidence interval-upper bound (percentage): 38.7. Not applicable; Estimated percentage: 2.3; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 11.4. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14v. Complex or unclear ballot; Major factor; Estimated percentage: 7.7; 95 percent confidence interval-lower bound (percentage): 3.0; 95 percent confidence interval-upper bound (percentage): 15.6. Minor factor; Estimated percentage: 17.2; 95 percent confidence interval-lower bound (percentage): 8.8; 95 percent confidence interval-upper bound (percentage): 28.8. Not a factor; Estimated percentage: 72.7; 95 percent confidence interval-lower bound (percentage): 60.3; 95 percent confidence interval-upper bound (percentage): 83.0. Not applicable; Estimated percentage: 2.5; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 10.7. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14w. Large number of non-English speaking voters; Major factor; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Minor factor; Estimated percentage: 14.3; 95 percent confidence interval-lower bound (percentage): 6.9; 95 percent confidence interval-upper bound (percentage): 25.3. Not a factor; Estimated percentage: 83.4; 95 percent confidence interval-lower bound (percentage): 72.2; 95 percent confidence interval-upper bound (percentage): 91.5. Not applicable; Estimated percentage: 2.3; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 8.5. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14x. Voter identification requirements; Major factor; Estimated percentage: 7.6; 95 percent confidence interval-lower bound (percentage): 2.3; 95 percent confidence interval-upper bound (percentage): 17.3. Minor factor; Estimated percentage: 17.9; 95 percent confidence interval-lower bound (percentage): 9.5; 95 percent confidence interval-upper bound (percentage): 29.2. Not a factor; Estimated percentage: 69.4; 95 percent confidence interval-lower bound (percentage): 56.8; 95 percent confidence interval-upper bound (percentage): 80.2. Not applicable; Estimated percentage: 5.2; 95 percent confidence interval-lower bound (percentage): 1.6; 95 percent confidence interval-upper bound (percentage): 12.1. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14y. Processing provisional voters; Major factor; Estimated percentage: 12.5; 95 percent confidence interval-lower bound (percentage): 5.7; 95 percent confidence interval-upper bound (percentage): 22.9. Minor factor; Estimated percentage: 31.6; 95 percent confidence interval-lower bound (percentage): 20.7; 95 percent confidence interval-upper bound (percentage): 44.1. Not a factor; Estimated percentage: 53.0; 95 percent confidence interval-lower bound (percentage): 40.8; 95 percent confidence interval-upper bound (percentage): 65.1. Not applicable; Estimated percentage: 3.0; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 9.0. No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.8. Response: 14z. Other (please specify below); Major factor; Estimated percentage: ; 95 percent confidence interval-lower bound (percentage): 8.8; 95 percent confidence interval-upper bound (percentage): 31.4. Minor factor; Estimated percentage: 5.2; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 14.6. Not a factor; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Not applicable; Estimated percentage: 21.1; 95 percent confidence interval-lower bound (percentage): 10.3; 95 percent confidence interval-upper bound (percentage): 35.9. No response; Estimated percentage: 22.8; 95 percent confidence interval-lower bound (percentage): 12.1; 95 percent confidence interval-upper bound (percentage): 36.9. Note: N/R indicates that we are not reporting results because the width of the confidence interval is greater than plus or minus 15 percentage points and the results are considered not reliable. [End of table] What other aspect was a factor? [open ended] 15. Does your jurisdiction have a formal goal for the maximum time that a voter should wait to begin to fill out a ballot? (Please check one response): Response: Yes (Go to question 16); Estimated percentage: 13.5; 95 percent confidence interval-lower bound (percentage): 9.8; 95 percent confidence interval-upper bound (percentage): 17.9. Response: No (Skip to question 17); Estimated percentage: 80.3; 95 percent confidence interval-lower bound (percentage): 75.7; 95 percent confidence interval-upper bound (percentage): 84.9. Response: Don't know (Skip to question 17); Estimated percentage: 6.3; 95 percent confidence interval-lower bound (percentage): 3.7; 95 percent confidence interval-upper bound (percentage): 9.8. [End of table] 16. If Yes: What is the maximum wait time goal? (Please check one response): Response: 10 minutes or less; Estimated: percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: 11 to 20 minutes; Estimated: percentage: 8.2; 95 percent confidence interval-lower bound (percentage): 2.0; 95 percent confidence interval-upper bound (percentage): 20.8. Response: 21 to 30 minutes; Estimated: percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: 31 to 60 minutes; Estimated: percentage: 3.8; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 13.3. Response: More than 60 minutes; Estimated: percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.0. Response: Don't know; Estimated: percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.0. Note: N/R indicates that we are not reporting results because the width of the confidence interval is greater than plus or minus 15 percentage points and the results are considered not reliable. [End of table] 17. At about how many polling places did wait times of greater than 60 minutes occur at any time on either the November 2012 or November 2008 General Election Day? (Please check one response): 2012: Response: No polling places; Estimated percentage: 79.5; 95 percent confidence interval-lower bound (percentage): 74.9; 95 percent confidence interval-upper bound (percentage): 84.1. Response: Only a few polling places; Estimated percentage: 8.9; 95 percent confidence interval-lower bound (percentage): 5.9; 95 percent confidence interval-upper bound (percentage): 12.9. Response: Less than half, but more than a few polling places; Estimated percentage: 0.9; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 2.6. Response: About half of all polling places; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 3.0. Response: More than half of all polling places; Estimated percentage: 0.6; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.4. Response: Don't know or don't remember; Estimated percentage: 8.0; 95 percent confidence interval-lower bound (percentage): 5.2; 95 percent confidence interval-upper bound (percentage): 11.8. Response: No response; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 2.8. [End of table] 2008: Response: No polling places; Estimated percentage: 68.2; 95 percent confidence interval-lower bound (percentage): 62.9; 95 percent confidence interval-upper bound (percentage): 73.5. Response: Only a few polling places; Estimated percentage: 8.1; 95 percent confidence interval-lower bound (percentage): 5.2; 95 percent confidence interval-upper bound (percentage): 11.9. Response: Less than half, but more than a few polling places; Estimated percentage: 1.3; 95 percent confidence interval-lower bound (percentage): 0.3; 95 percent confidence interval-upper bound (percentage): 3.3. Response: About half of all polling places; Estimated percentage: 1.5; 95 percent confidence interval-lower bound (percentage): 0.4; 95 percent confidence interval-upper bound (percentage): 3.8. Response: More than half of all polling places; Estimated percentage: 0.7; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.4. Response: Don't know or don't remember; Estimated percentage: 18.3; 95 percent confidence interval-lower bound (percentage): 13.7; 95 percent confidence interval-upper bound (percentage): 22.9. Response: No response; Estimated percentage: 1.9; 95 percent confidence interval-lower bound (percentage): 0.7; 95 percent confidence interval-upper bound (percentage): 4.3. [End of table] 18. On the November 2012 General Election Day, did any voters have to wait in line to turn in their ballot to a machine, poll worker, or ballot box after completing their ballot? We understand that for DRE machines, voters submit their ballots immediately after completing them, but for other methods, such as optical/digital scan or paper ballots, voters may have to wait in line to turn in their ballot to feed through a machine or submit their ballot to a poll worker or ballot box. (Please check one response): Response: Yes (Go to question 19); Estimated percentage: 15.5; 95 percent confidence interval-lower bound (percentage): 11.7; 95 percent confidence interval-upper bound (percentage): 20.0. Response: No (Skip to question 21); Estimated percentage: 71.5; 95 percent confidence interval-lower bound (percentage): 66.4; 95 percent confidence interval-upper bound (percentage): 76.7. Response: Don't know (Skip to question 21); Estimated percentage: 13.0; 95 percent confidence interval-lower bound (percentage): 9.2; 95 percent confidence interval-upper bound (percentage): 17.6. [End of table] 19. If yes: On average, for all of the polling places in your jurisdiction for the November 2012 General Election Day, how long did voters typically have to wait in line to turn in their ballots for counting after completing them? (Please check one response): Response: 5 minutes or less; Estimated percentage: 89.9; 95 percent confidence interval-lower bound (percentage): 77.0; 95 percent confidence interval-upper bound (percentage): 96.9. Response: 6 to 15 minutes; Estimated percentage: 6.0; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 17.3. Response: 16 to 30 minutes; Estimated percentage: 3.8; 95 percent confidence interval-lower bound (percentage): 0.3; 95 percent confidence interval-upper bound (percentage): 15.3. Response: More than 30 minutes; Estimated percentage: 0.3; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 9.4. Response: Don't know; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 5.1. [End of table] 20. Which of the following, if any, were reasons voters had to wait to turn in ballots for counting after voters completed their ballots? (Please check one response for each row): 20a.Voters periodically had to wait in short lines to turn in ballots, but lines were minimal and not due to a specific reason; Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: 12.0; 95 percent confidence interval-lower bound (percentage): 3.8; 95 percent confidence interval-upper bound (percentage): 26.4. Response: No response; Estimated percentage: 3.7; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 16.1. 20b. Not enough counting machines (e.g., optical/digital scan); Response: Yes; Estimated percentage: 9.2; 95 percent confidence interval-lower bound (percentage): 2.2; 95 percent confidence interval-upper bound (percentage): 23.7. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 3.5; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 16.9. 20c. Counting machine failure (e.g., optical/digital scan); Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. 20d. Long ballot to feed through machine; Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 3.2; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 15.8. 20e. Design/layout of polling places; Response: Yes; Estimated percentage: 6.3; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 17.7. Response: No; Estimated percentage: 90.2; 95 percent confidence interval-lower bound (percentage): 76.5; 95 percent confidence interval-upper bound (percentage): 97.4. Response: No response; Estimated percentage: 3.5; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 17.1. 20f. Other (Please specify below); Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Note: N/R indicates that we are not reporting results because the width of the confidence interval is greater than plus or minus 15 percentage points and the results are considered not reliable. [End of table] What was the other reason that voters had to wait to turn in ballots for counting? [open ended] Part III: Policies and Practices in Your Jurisdiction: 21. Did your jurisdiction use any of the following policies or practices for the November 2012 general election? (Please check one response for each row): 21a. Mail-in voting (e.g., absentee); Response: Used; Estimated percentage: 98.4; 95 percent confidence interval-lower bound (percentage): 96.3; 95 percent confidence interval-upper bound (percentage): 99.5. Response: Didn't use; Estimated percentage: 1.6; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 3.7. Response: Don't know; Estimated percentage: 0.1; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.7. Response: No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. 21b. In-person voting prior to Election Day (e.g., early or in-person absentee voting); Response: Used; Estimated percentage: 88.6; 95 percent confidence interval-lower bound (percentage): 84.5; 95 percent confidence interval-upper bound (percentage): 92.0. Response: Didn't use; Estimated percentage: 11.4; 95 percent confidence interval-lower bound (percentage): 8.0; 95 percent confidence interval-upper bound (percentage): 15.5. Response: Don't know; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. Response: No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. 21c. Vote centers on Election Day (polling places where any voter in the jurisdiction can vote regardless of precinct); Response: Used; Estimated percentage: 5.2; 95 percent confidence interval-lower bound (percentage): 2.9; 95 percent confidence interval-upper bound (percentage): 8.4. Response: Didn't use; Estimated percentage: 92.9; 95 percent confidence interval-lower bound (percentage): 89.2; 95 percent confidence interval-upper bound (percentage): 95.6. Response: Don't know; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. Response: No response; Estimated percentage: 1.9; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 4.4. 21d. Election Day registration; Response: Used; Estimated percentage: 19.4; 95 percent confidence interval-lower bound (percentage): 14.8; 95 percent confidence interval-upper bound (percentage): 24.1. Response: Didn't use; Estimated percentage: 76.9; 95 percent confidence interval-lower bound (percentage): 72.0; 95 percent confidence interval-upper bound (percentage): 81.9. Response: Don't know; Estimated percentage: 0.8; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.7. Response: No response; Estimated percentage: 2.8; 95 percent confidence interval-lower bound (percentage): 1.2; 95 percent confidence interval-upper bound (percentage): 5.7. 21e. Electronic poll books at polling places on Election Day; Response: Used; Estimated percentage: 29.1; 95 percent confidence interval-lower bound (percentage): 23.7; 95 percent confidence interval-upper bound (percentage): 34.5. Response: Didn't use; Estimated percentage: 70.4; 95 percent confidence interval-lower bound (percentage): 65.0; 95 percent confidence interval-upper bound (percentage): 75.8. Response: Don't know; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. Response: No response; Estimated percentage: 0.5; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.4. 21f. Paper poll books at polling places on Election Day; Response: Used; Estimated percentage: 77.2; 95 percent confidence interval-lower bound (percentage): 72.2; 95 percent confidence interval-upper bound (percentage): 82.1. Response: Didn't use; Estimated percentage: 21.9; 95 percent confidence interval-lower bound (percentage): 17.0; 95 percent confidence interval-upper bound (percentage): 26.8. Response: Don't know; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. Response: No response; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 3.2. 21g. Ballots available in different languages; Response: Used; Estimated percentage: 16.6; 95 percent confidence interval-lower bound (percentage): 12.6; 95 percent confidence interval-upper bound (percentage): 21.2. Response: Didn't use; Estimated percentage: 80.4; 95 percent confidence interval-lower bound (percentage): 75.9; 95 percent confidence interval-upper bound (percentage): 84.9. Response: Don't know; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 4.3. Response: No response; Estimated percentage: 1.3; 95 percent confidence interval-lower bound (percentage): 0.3; 95 percent confidence interval-upper bound (percentage): 3.5. 21h. Standardized training for poll workers; Response: Used; Estimated percentage: 97.6; 95 percent confidence interval-lower bound (percentage): 95.0; 95 percent confidence interval-upper bound (percentage): 99.1. Response: Didn't use; Estimated percentage: 1.5; 95 percent confidence interval-lower bound (percentage): 0.5; 95 percent confidence interval-upper bound (percentage): 3.8. Response: Don't know; Estimated percentage: 0.9; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 2.9. Response: No response; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 0.9. 21i. Other (please specify below); Response: Used; Estimated percentage: 8.0; 95 percent confidence interval-lower bound (percentage): 4.1; 95 percent confidence interval-upper bound (percentage): 13.7. Response: Didn't use; Estimated percentage: 27.8; 95 percent confidence interval-lower bound (percentage): 20.1; 95 percent confidence interval-upper bound (percentage): 35.4. Response: Don't know; Estimated percentage: 5.5; 95 percent confidence interval-lower bound (percentage): 2.2; 95 percent confidence interval-upper bound (percentage): 11.1. Response: No response; Estimated percentage: 58.7; 95 percent confidence interval-lower bound (percentage): 50.4; 95 percent confidence interval-upper bound (percentage): 67.1. [End of table] What other policies or practices did your jurisdiction use? [open ended] 22. Did your jurisdiction conduct a formal audit or investigation of the possible causes of long voter wait times on the November 2012 General Election Day? (Please check one response): Response: Yes; Estimated percentage: 4.7; 95 percent confidence interval-lower bound (percentage): 2.6; 95 percent confidence interval-upper bound (percentage): 7.7. Response: No; Estimated percentage: 94.4; 95 percent confidence interval-lower bound (percentage): 91.1; 95 percent confidence interval-upper bound (percentage): 96.7. Response: Don't know; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 3.1. [End of table] 23. Has your jurisdiction revised or is it in the process of revising any of its Election Day policies or procedures since the November 2012 general election specifically to address any of the possible causes of long voter wait times? (Please check one response): Response: Yes (Go to question 24); Estimated percentage: 18.1; 95 percent confidence interval-lower bound (percentage): 13.7; 95 percent confidence interval-upper bound (percentage): 22.4. Response: No (Skip to question 25); Estimated percentage: 78.2; 95 percent confidence interval-lower bound (percentage): 73.4; 95 percent confidence interval-upper bound (percentage): 82.9. Response: Don't know (Skip to question 25); Estimated percentage: 3.8; 95 percent confidence interval-lower bound (percentage): 1.9; 95 percent confidence interval-upper bound (percentage): 6.7. [End of table] 24. If yes: which policies or procedures were revised or are in the process of being revised specifically to address any of the possible causes of long voter wait times on the November 2012 General Election Day? (Please check one response on each row): 24a. Options for voting outside of Election Day (e.g., early or mail- in voting); Response: Yes; Estimated percentage: 30.5; 95 percent confidence interval-lower bound (percentage): 18.2; 95 percent confidence interval-upper bound (percentage): 45.1. Response: No; Estimated percentage: 59.6; 95 percent confidence interval-lower bound (percentage): 46.0; 95 percent confidence interval-upper bound (percentage): 73.2. Response: No response; Estimated percentage: 10.0; 95 percent confidence interval-lower bound (percentage): 3.3; 95 percent confidence interval-upper bound (percentage): 22.0. 24b. Number of locations for in-person voting prior to Election Day; Response: Yes; Estimated percentage: 23.8; 95 percent confidence interval-lower bound (percentage): 12.8; 95 percent confidence interval-upper bound (percentage): 38.1. Response: No; Estimated percentage: 68.3; 95 percent confidence interval-lower bound (percentage): 53.4; 95 percent confidence interval-upper bound (percentage): 80.9. Response: No response; Estimated percentage: 7.9; 95 percent confidence interval-lower bound (percentage): 2.1; 95 percent confidence interval-upper bound (percentage): 19.6. 24c. Number of days or hours for in-person voting prior to Election Day; Response: Yes; Estimated percentage: 23.6; 95 percent confidence interval-lower bound (percentage): 13.0; 95 percent confidence interval-upper bound (percentage): 37.3. Response: No; Estimated percentage: 70.6; 95 percent confidence interval-lower bound (percentage): 56.1; 95 percent confidence interval-upper bound (percentage): 82.6. Response: No response; Estimated percentage: 5.8; 95 percent confidence interval-lower bound (percentage): 1.1; 95 percent confidence interval-upper bound (percentage): 16.7. 24d. Number of Election Day polling places; Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 4.6; 95 percent confidence interval-lower bound (percentage): 0.7; 95 percent confidence interval-upper bound (percentage): 14.8. 24e. Creation of vote centers (polling places where any voter in the jurisdiction can vote regardless of precinct); Response: Yes; Estimated percentage: 12.2; 95 percent confidence interval-lower bound (percentage): 4.9; 95 percent confidence interval-upper bound (percentage): 23.8. Response: No; Estimated percentage: 79.9; 95 percent confidence interval-lower bound (percentage): 66.3; 95 percent confidence interval-upper bound (percentage): 89.8. Response: No response; Estimated percentage: 8.0; 95 percent confidence interval-lower bound (percentage): 2.1; 95 percent confidence interval-upper bound (percentage): 19.7. 24f. Design/layout of polling places; Response: Yes; Estimated percentage: 64.6; 95 percent confidence interval-lower bound (percentage): 49.8; 95 percent confidence interval-upper bound (percentage): 77.6. Response: No; Estimated percentage: 31.1; 95 percent confidence interval-lower bound (percentage): 18.8; 95 percent confidence interval-upper bound (percentage): 45.7. Response: No response; Estimated percentage: 4.4; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 14.1. 24g. Number of voting machines; Response: Yes; Estimated percentage: 29.4; 95 percent confidence interval-lower bound (percentage): 17.5; 95 percent confidence interval-upper bound (percentage): 43.9. Response: No; Estimated percentage: 63.8; 95 percent confidence interval-lower bound (percentage): 48.9; 95 percent confidence interval-upper bound (percentage): 76.9. Response: No response; Estimated percentage: 6.8; 95 percent confidence interval-lower bound (percentage): 1.6; 95 percent confidence interval-upper bound (percentage): 17.9. 24h. Voting machine testing; Response: Yes; Estimated percentage: 26.0; 95 percent confidence interval-lower bound (percentage): 14.6; 95 percent confidence interval-upper bound (percentage): 40.5. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 10.3; 95 percent confidence interval-lower bound (percentage): 3.2; 95 percent confidence interval-upper bound (percentage): 23.1. 24i. Type of voting method or machine; Response: Yes; Estimated percentage: 18.0; 95 percent confidence interval-lower bound (percentage): 8.6; 95 percent confidence interval-upper bound (percentage): 31.4. Response: No; Estimated percentage: 75.2; 95 percent confidence interval-lower bound (percentage): 61.0; 95 percent confidence interval-upper bound (percentage): 86.3. Response: No response; Estimated percentage: 6.8; 95 percent confidence interval-lower bound (percentage): 1.6; 95 percent confidence interval-upper bound (percentage): 17.9. 24j. Number of poll workers at polling places; Response: Yes; Estimated percentage: 67.1; 95 percent confidence interval-lower bound (percentage): 52.4; 95 percent confidence interval-upper bound (percentage): 79.7. Response: No; Estimated percentage: 30.7; 95 percent confidence interval-lower bound (percentage): 18.5; 95 percent confidence interval-upper bound (percentage): 45.3. Response: No response; Estimated percentage: 2.2; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 10.9. 24k. Training of poll workers; Response: Yes; Estimated percentage: 74.4; 95 percent confidence interval-lower bound (percentage): 59.9; 95 percent confidence interval-upper bound (percentage): 85.7. Response: No; Estimated percentage: 23.4; 95 percent confidence interval-lower bound (percentage): 12.6; 95 percent confidence interval-upper bound (percentage): 37.6. Response: No response; Estimated percentage: 2.2; 95 percent confidence interval-lower bound (percentage): 0.1; 95 percent confidence interval-upper bound (percentage): 10.9. 24l. Revised voter turnout estimation procedures/methods; Response: Yes; Estimated percentage: 33.1; 95 percent confidence interval-lower bound (percentage): 20.7; 95 percent confidence interval-upper bound (percentage): 47.6. Response: No; Estimated percentage: 57.8; 95 percent confidence interval-lower bound (percentage): 44.3; 95 percent confidence interval-upper bound (percentage): 71.4. Response: No response; Estimated percentage: 9.1; 95 percent confidence interval-lower bound (percentage): 2.9; 95 percent confidence interval-upper bound (percentage): 20.4. 24m. Redraw precinct boundaries; Response: Yes; Estimated percentage: 19.1; 95 percent confidence interval-lower bound (percentage): 9.1; 95 percent confidence interval-upper bound (percentage): 33.2. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 10.0; 95 percent confidence interval-lower bound (percentage): 3.1; 95 percent confidence interval-upper bound (percentage): 22.5. 24n. Consolidation/changes in polling places; Response: Yes; Estimated percentage: 41.7; 95 percent confidence interval-lower bound (percentage): 27.9; 95 percent confidence interval-upper bound (percentage): 55.5. Response: No; Estimated percentage: 48.4; 95 percent confidence interval-lower bound (percentage): 34.4; 95 percent confidence interval-upper bound (percentage): 62.4. Response: No response; Estimated percentage: 9.9; 95 percent confidence interval-lower bound (percentage): 3.0; 95 percent confidence interval-upper bound (percentage): 22.6. 24o. Use of paper poll books; Response: Yes; Estimated percentage: 23.7; 95 percent confidence interval-lower bound (percentage): 13.1; 95 percent confidence interval-upper bound (percentage): 37.4. Response: No; Estimated percentage: 71.6; 95 percent confidence interval-lower bound (percentage): 57.5; 95 percent confidence interval-upper bound (percentage): 83.2. Response: No response; Estimated percentage: 4.7; 95 percent confidence interval-lower bound (percentage): 0.8; 95 percent confidence interval-upper bound (percentage): 14.5. 24p. Use of electronic poll books; Response: Yes; Estimated percentage: 41.4; 95 percent confidence interval-lower bound (percentage): 27.8; 95 percent confidence interval-upper bound (percentage): 55.0. Response: No; Estimated percentage: 50.8; 95 percent confidence interval-lower bound (percentage): 37.0; 95 percent confidence interval-upper bound (percentage): 64.7. Response: No response; Estimated percentage: 7.8; 95 percent confidence interval-lower bound (percentage): 2.2; 95 percent confidence interval-upper bound (percentage): 18.7. 24q. Improving accuracy of voter registration rolls; Response: Yes; Estimated percentage: 39.3; 95 percent confidence interval-lower bound (percentage): 25.8; 95 percent confidence interval-upper bound (percentage): 54.0. Response: No; Estimated percentage: 50.9; 95 percent confidence interval-lower bound (percentage): 37.0; 95 percent confidence interval-upper bound (percentage): 64.9. Response: No response; Estimated percentage: 9.8; 95 percent confidence interval-lower bound (percentage): 3.4; 95 percent confidence interval-upper bound (percentage): 21.1. 24r. Election Day voter registration; Response: Yes; Estimated percentage: 7.5; 95 percent confidence interval-lower bound (percentage): 2.0; 95 percent confidence interval-upper bound (percentage): 18.4. Response: No; Estimated percentage: 82.7; 95 percent confidence interval-lower bound (percentage): 69.3; 95 percent confidence interval-upper bound (percentage): 91.9. Response: No response; Estimated percentage: 9.8; 95 percent confidence interval-lower bound (percentage): 3.2; 95 percent confidence interval-upper bound (percentage): 21.6. 24s. Ballot simplification; Response: Yes; Estimated percentage: 10.9; 95 percent confidence interval-lower bound (percentage): 4.2; 95 percent confidence interval-upper bound (percentage): 21.9. Response: No; Estimated percentage: 74.1; 95 percent confidence interval-lower bound (percentage): 59.6; 95 percent confidence interval-upper bound (percentage): 85.5. Response: No response; Estimated percentage: 15.0; 95 percent confidence interval-lower bound (percentage): 6.2; 95 percent confidence interval-upper bound (percentage): 28.8. 24t. Accommodation of non-English speaking voters; Response: Yes; Estimated percentage: 14.7; 95 percent confidence interval-lower bound (percentage): 6.9; 95 percent confidence interval-upper bound (percentage): 26.1. Response: No; Estimated percentage: 64.4; 95 percent confidence interval-lower bound (percentage): 49.6; 95 percent confidence interval-upper bound (percentage): 77.4. Response: No response; Estimated percentage: 20.9; 95 percent confidence interval-lower bound (percentage): 10.4; 95 percent confidence interval-upper bound (percentage): 35.4. 24u. Voter identification requirements; Response: Yes; Estimated percentage: 21.5; 95 percent confidence interval-lower bound (percentage): 10.8; 95 percent confidence interval-upper bound (percentage): 36.1. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 15.4; 95 percent confidence interval-lower bound (percentage): 6.5; 95 percent confidence interval-upper bound (percentage): 29.0. 24v. Revised provisional vote procedures; Response: Yes; Estimated percentage: 22.9; 95 percent confidence interval-lower bound (percentage): 12.1; 95 percent confidence interval-upper bound (percentage): 37.1. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: 10.4; 95 percent confidence interval-lower bound (percentage): 3.6; 95 percent confidence interval-upper bound (percentage): 22.2. 24w. Others (please specify below); Response: Yes; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence interval-lower bound (percentage): N/R; 95 percent confidence interval-upper bound (percentage): N/R. Note: N/R indicates that we are not reporting results because the width of the confidence interval is greater than plus or minus 15 percentage points and the results are considered not reliable. [End of table] What other policies or procedures were revised or are in the process of being revised? [open ended] 25. What policies and procedures are most important to minimize or reduce voter wait time in your jurisdiction? Please answer this question whether or not your jurisdiction has experienced long voter wait times. [open ended] 26. Which of the following best describes the resources that were available to your jurisdiction for the November 2012 General Election Day? (Please check one response): Response: There were enough resources to comfortably conduct Election Day operations (Skip to question 28); Estimated: percentage: 80.9; 95 percent confidence interval-lower bound (percentage): 76.5; 95 percent confidence interval-upper bound (percentage): 85.3. Response: Resources were tight, but Election Day operations were conducted as planned (Skip to question 28); Estimated: percentage: 17.2; 95 percent confidence interval-lower bound (percentage): 13.2; 95 percent confidence interval-upper bound (percentage): 22.0. Response: There were resource shortages and some Election Day operations were impacted by these shortages (Go to question 27); Estimated: percentage: 1.9; 95 percent confidence interval-lower bound (percentage): 0.7; 95 percent confidence interval-upper bound (percentage): 4.0. [End of table] 27. Which of the following activities, if any, were impacted by the availability of resources? (Check all that apply): 27a. The number of polling locations used on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27b. The type (optimal size and configuration) of polling locations used on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27c. The number of poll workers used on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27d. The training of poll workers; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27e. The number of voting machines used on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27f. The type of voting machine/technology used on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27g. The availability of other technology, for instance electronic poll books; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27h. Voter education efforts before Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27i. Voter education efforts on Election Day; Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. 27j. Other (please specify below); Response: Yes; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Response: No response; Estimated percentage: N/R; 95 percent confidence Interval-lower bound (percentage): N/R; 95 percent confidence Interval-upper bound (percentage): N/R. Note: N/R indicates that we are not reporting results because the width of the confidence interval is greater than plus or minus 15 percentage points and the results are considered not reliable. [End of table] What other activities were impacted? [open ended] Part IV: 2012 General Election Characteristics: For the November 2012 General Election Day, which types of voting methods were used? For those that were used, please provide the numbers of machines and/or ballots cast. (Please check at least one response on each row): Number of machines: Number of ballots cast: 28a. Electronic (direct recording electronic-DRE); Response: Used; Estimated percentage: 51.3; 95 percent confidence interval-lower bound (percentage): 45.4; 95 percent confidence interval-upper bound (percentage): 57.1. Response: Not used; Estimated percentage: 46.9; 95 percent confidence interval-lower bound (percentage): 41.0; 95 percent confidence interval-upper bound (percentage): 52.8. Response: No response; Estimated percentage: 1.8; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 4.4. 28b. Optical/digital scan; Response: Used; Estimated percentage: 72.9; 95 percent confidence interval-lower bound (percentage): 67.6; 95 percent confidence interval-upper bound (percentage): 78.1. Response: Not used; Estimated percentage: 24.5; 95 percent confidence interval-lower bound (percentage): 19.5; 95 percent confidence interval-upper bound (percentage): 29.6. Response: No response; Estimated percentage: 2.6; 95 percent confidence interval-lower bound (percentage): 1.0; 95 percent confidence interval-upper bound (percentage): 5.3. 28c. Lever machine; Response: Used; Estimated percentage: 0; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 1.0. Response: Not used; Estimated percentage: 98.0; 95 percent confidence interval-lower bound (percentage): 95.3; 95 percent confidence interval-upper bound (percentage): 99.4. Response: No response; Estimated percentage: 2.0; 95 percent confidence interval-lower bound (percentage): 0.6; 95 percent confidence interval-upper bound (percentage): 4.7. 28d. Punch card ballot; Response: Used; Estimated percentage: 0.5; 95 percent confidence interval-lower bound (percentage): 0; 95 percent confidence interval-upper bound (percentage): 2.7. Response: Not used; Estimated percentage: 98.5; 95 percent confidence interval-lower bound (percentage): 96.1; 95 percent confidence interval-upper bound (percentage): 99.6. Response: No response; Estimated percentage: 1.0; 95 percent confidence interval-lower bound (percentage): 0.2; 95 percent confidence interval-upper bound (percentage): 3.0. 28e. Paper (hand-counted) ballot; Response: Used; Estimated percentage: 18.1; 95 percent confidence interval-lower bound (percentage): 13.5; 95 percent confidence interval-upper bound (percentage): 23.5. Response: Not used; Estimated percentage: 76.3; 95 percent confidence interval-lower bound (percentage): 70.9; 95 percent confidence interval-upper bound (percentage): 81.7. Response: No response; Estimated percentage: 5.6; 95 percent confidence interval-lower bound (percentage): 3.0; 95 percent confidence interval-upper bound (percentage): 9.5. 28f. Other (please specify below); Response: Used; Estimated percentage: 11.1; 95 percent confidence interval-lower bound (percentage): 6.9; 95 percent confidence interval-upper bound (percentage): 16.6. Response: Not used; Estimated percentage: 56.4; 95 percent confidence interval-lower bound (percentage): 49.2; 95 percent confidence interval-upper bound (percentage): 63.5. Response: No response; Estimated percentage: 32.5; 95 percent confidence interval-lower bound (percentage): 25.8; 95 percent confidence interval-upper bound (percentage): 39.3. [End of table] What other type of voting method was used? [open ended] Number of ballots cast: 29. For the November 2012 general election, how many votes were cast through the following methods? [open ended] In-person voting on Election Day at a polling place (excluding provisional voting): Provisional voting on Election Day at a polling place (both accepted and rejected ballots): All voting that occurred outside of Election Day (e.g., any type of early, absentee, and mail-in ballots, including mail-in ballots that were submitted in person on Election Day): 30. On the November 2012 General Election Day, how many poll workers were used in your jurisdiction? By poll workers, we mean those individuals recruited specifically for the purpose of working at polling places on Election Day. [open ended] Number of poll workers: 31. How many hours of election administration training did poll workers in your jurisdiction receive in preparation for the November 2012 general election? Providing an estimate is fine. [open ended] Hours of election administration training for typical first-time poll worker: Hours of election administration training for typical returning poll worker: Hours of election administration training for typical polling place supervisor or presiding judge: 32. What was the total number of ballot questions (propositions) and elected offices (races) that your jurisdiction was asked to put on applicable ballots in the November 2012 general election? [open ended] Ballot questions (propositions): Elected offices (races): 33. How many pages or screens was an average ballot in your jurisdiction? [open ended] pages: screens: 34. Which of the following information, if any, did your jurisdiction provide to educate the public prior to the November 2012 general election? (Please check at least one response in each row): 34a. Specific polling place location information; Response: On website; Estimated percentage: 78.2; 95 percent confidence interval-lower bound (percentage): 73.5; 95 percent confidence interval-upper bound (percentage): 82.9. Response: By mail; Estimated percentage: 22.4; 95 percent confidence interval-lower bound (percentage): 17.7; 95 percent confidence interval-upper bound (percentage): 27.0. Response: Other method(s); Estimated percentage: 60.3; 95 percent confidence interval-lower bound (percentage): 54.7; 95 percent confidence interval-upper bound (percentage): 65.9. Response: Did not provide; Estimated percentage: 1.3; 95 percent confidence interval-lower bound (percentage): 0.3; 95 percent confidence interval-upper bound (percentage): 3.6. 34b. Sample ballots; Response: On website; Estimated percentage: 63.0; 95 percent confidence interval-lower bound (percentage): 57.4; 95 percent confidence interval-upper bound (percentage): 68.5. Response: By mail; Estimated percentage: 16.5; 95 percent confidence interval-lower bound (percentage): 12.5; 95 percent confidence interval-upper bound (percentage): 21.3. Response: Other method(s); Estimated percentage: 67.2; 95 percent confidence interval-lower bound (percentage): 61.9; 95 percent confidence interval-upper bound (percentage): 72.5. Response: Did not provide; Estimated percentage: 2.4; 95 percent confidence interval-lower bound (percentage): 0.9; 95 percent confidence interval-upper bound (percentage): 5.0. 34c. Information about their registration status; Response: On website; Estimated percentage: 48.2; 95 percent confidence interval-lower bound (percentage): 42.4; 95 percent confidence interval-upper bound (percentage): 53.9. Response: By mail; Estimated percentage: 20.8; 95 percent confidence interval-lower bound (percentage): 16.2; 95 percent confidence interval-upper bound (percentage): 25.5. Response: Other method(s); Estimated percentage: 41.0; 95 percent confidence interval-lower bound (percentage): 35.2; 95 percent confidence interval-upper bound (percentage): 46.7. Response: Did not provide; Estimated percentage: 15.5; 95 percent confidence interval-lower bound (percentage): 11.4; 95 percent confidence interval-upper bound (percentage): 20.3. 34d. Instructions about how to cast a vote using the jurisdiction's voting method; Response: On website; Estimated percentage: 34.5; 95 percent confidence interval-lower bound (percentage): 29.3; 95 percent confidence interval-upper bound (percentage): 39.8. Response: By mail; Estimated percentage: 11.8; 95 percent confidence interval-lower bound (percentage): 8.4; 95 percent confidence interval-upper bound (percentage): 15.9. Other method(s); Response: Estimated percentage: 65.2; 95 percent confidence interval-lower bound (percentage): 59.9; 95 percent confidence interval-upper bound (percentage): 70.5. Response: Did not provide; Estimated percentage: 11.4; 95 percent confidence interval-lower bound (percentage): 7.9; 95 percent confidence interval-upper bound (percentage): 15.7. 34e. Information about options to vote outside of Election Day (e.g., early or mail-in voting); Response: On website; Estimated percentage: 62.7; 95 percent confidence interval-lower bound (percentage): 57.1; 95 percent confidence interval-upper bound (percentage): 68.3. Response: By mail; Estimated percentage: 18.4; 95 percent confidence interval-lower bound (percentage): 14.0; 95 percent confidence interval-upper bound (percentage): 22.7. Response: Other method(s); Estimated percentage: 65.4; 95 percent confidence interval-lower bound (percentage): 60.0; 95 percent confidence interval-upper bound (percentage): 70.9. Response: Did not provide; Estimated percentage: 7.9; 95 percent confidence interval-lower bound (percentage): 5.0; 95 percent confidence interval-upper bound (percentage): 11.6. 34f. Full text statements by candidates or information about ballot questions; Response: On website; Estimated percentage: 19.6; 95 percent confidence interval-lower bound (percentage): 15.1; 95 percent confidence interval-upper bound (percentage): 24.2. Response: By mail; Estimated percentage: 11.1; 95 percent confidence interval-lower bound (percentage): 7.8; 95 percent confidence interval-upper bound (percentage): 15.2. Response: Other method(s); Estimated percentage: 29.3; 95 percent confidence interval-lower bound (percentage): 24.0; 95 percent confidence interval-upper bound (percentage): 34.6. Response: Did not provide; Estimated percentage: 49.1; 95 percent confidence interval-lower bound (percentage): 43.2; 95 percent confidence interval-upper bound (percentage): 55.0. 34g. Information for voters in a language other than English; Response: On website; Estimated percentage: 13.4; 95 percent confidence interval-lower bound (percentage): 9.9; 95 percent confidence interval-upper bound (percentage): 17.6. Response: By mail; Estimated percentage: 6.3; 95 percent confidence interval-lower bound (percentage): 4.0; 95 percent confidence interval-upper bound (percentage): 9.5. Response: Other method(s); Estimated percentage: 20.0; 95 percent confidence interval-lower bound (percentage): 15.4; 95 percent confidence interval-upper bound (percentage): 24.6. Response: Did not provide; Estimated percentage: 61.6; 95 percent confidence interval-lower bound (percentage): 55.9; 95 percent confidence interval-upper bound (percentage): 67.2. 34h. Information specific to voters with disabilities; Response: On website; Estimated percentage: 32.7; 95 percent confidence interval-lower bound (percentage): 27.3; 95 percent confidence interval-upper bound (percentage): 38.1. Response: By mail; Estimated percentage: 11.3; 95 percent confidence interval-lower bound (percentage): 8.0; 95 percent confidence interval-upper bound (percentage): 15.4. Response: Other method(s); Estimated percentage: 57.3; 95 percent confidence interval-lower bound (percentage): 51.6; 95 percent confidence interval-upper bound (percentage): 62.9. Response: Did not provide; Estimated percentage: 17.7; 95 percent confidence interval-lower bound (percentage): 13.5; 95 percent confidence interval-upper bound (percentage): 22.6. [End of table] 35. What, if anything, could the federal government do to help address long voter wait times? [open ended] Part V: Other Comments: 36. Do you have any other comments you feel are important about Election Day 2012 processes that were not included above that may be related to voter wait times? [open ended] [End of section] Appendix III: Estimates of Average Wait Times by State on Election Day 2012 Based on Nationwide Public Opinion Survey: The Cooperative Congressional Election Study (CCES) has been conducted since 2006 to study congressional elections and representation using large-scale national surveys. The 2012 CCES surveyed 54,535 U.S. citizens aged 18 and over by Internet about their views and experiences before and after Election Day 2012.[Footnote 99] Respondents who reported voting in person (either prior to Election Day or on Election Day) were asked about the length of time they recalled waiting in line to vote. Figure 7 shows estimated average wait times by state on Election Day 2012 based on the data collected through the 2012 CCES survey.[Footnote 100] Figure 7: Estimates of Average Wait Times by State on Election Day 2012 Based on Nationwide Public Opinion Survey: [Refer to PDF for image: vertical bar graph] Average wait time (in minutes) by state: Alaska: 1.4; Vermont: 1.4; South Dakota: 2.1; Wyoming: 2.7; Nebraska: 3.7; Delaware: 4.1; Maine: 4.2; Iowa: 4.5; New Jersey: 5.4; Minnesota: 6.3; Connecticut: 6.4; Illinois: 6.5; California: 6.5; Massachusetts: 6.9; Kentucky: 7.9; Arizona: 8.3; Pennsylvania: 8.3; Wisconsin: 8.5; Idaho: 8.7; Ohio: 9.0; North Carolina: 9.1; Alabama: 9.2; Missouri: 10.2; Kansas: 11.3; Texas: 11.6; New York: 12.1; Indiana: 12.2; Georgia: 13.7; Louisiana: 13.9; Tennessee: 14.6; Oklahoma: 16.8; Michigan: 18.8; South Carolina: 26.3; Virginia: 27.0; Maryland: 28.9; Florida: 34.1. Two states were excluded from the figure because they were vote-by- mail states, and 12 states and the District of Columbia were excluded because of the relative imprecision of their estimates.[A] Source: GAO analysis of data from the 2012 Cooperative Congressional Election Study. GAO-14-850. Notes: We replicated a method used by Charles Stewart III when analyzing CCES data in "Waiting to Vote in 2012," Journal of Law and Politics. Specifically, we calculated average wait times by recoding response categories to the midpoint of the category. For example, the "none at all" response category was recoded as 0 minutes, and the "1- 10 minutes" response category was recoded as 5 minutes. (Respondents who waited more than an hour provided wait times in minutes). [A] Oregon and Washington were excluded because, as of the November 2012 election, they were vote-by-mail states. In addition to the District of Columbia, the 12 states with relatively imprecise estimates were Arkansas, Colorado, Hawaii, Mississippi, Montana, Nevada, New Hampshire, New Mexico, North Dakota, Rhode Island, Utah, and West Virginia. The estimated proportion of voters in these states who waited 10 minutes or less had a margin of error greater than plus or minus 10 percentage points. We excluded these states and the District of Columbia from the figure because of the relative imprecision of their estimates. [End of figure] [End of section] Appendix IV: Comments from the Supervisor of Elections for Lee County, Florida: Sharon L. Harrington: Supervisor of Elections: Lee County: Constitutional Complex: P.O, Box 2545: Fort Myers, Florida 33902: (239) Lee-Vote: (239) 533-8683: Fax (239) 533·6310: [hyperlink, http://www.leeelections.com] To: Johanna Wong: From: Sharon l. Harrington - Lee County Supervisor of Elections: Date: September 15, 2014: Subject: GAO Voter Wait Times Report: Dear Johanna, I sincerely believe that the information reported here reflects a true and accurate accounting of what transpired during the 2012 general election, As this report clearly shows, it wasn't just one issue that contributed to the long wait lines Lee County experienced, but a series of issues culminating in what we now refer to as "the perfect storm", I have been with our office for 25 years and have served as the elected Supervisor of Elections since 2004. During those 25 years, I have been involved in many situations that drastically changed the way we administer elections today. The passage of the National Voter Registration Act (NVRA) in the early 90's completely changed the way in which election administrators provided access to the voter registration process. Then we had the infamous 2000 election that resulted in the passage of the Help America Vote Act in 2002 that drastically change the methods of voting provided by each jurisdiction, In 2004 shortly after I was sworn in as Supervisor, we were hit by Hurricane Charley just a few weeks before the August primary. We lost a large number of our polling locations on the barrier islands and along the gulf shore and had to scramble to provide a place for voters to go on Election Day. This brought about the emergence of "early voting". In Charlotte County just north of us, they sustained more of a direct hit from Charley. They not only lost almost all of their polling locations, but their residents/voters were scattered everywhere across the county, It was during this time that a program was developed that would allow a voter to cast a ballot in a temporary "polling place" (most of them tents) regardless of where their original residence had been by printing the ballot style that they would have voted on initially. In 2006 we were mandated by new legislation to change our method of voting from DRE's, electronic touch screens to an optical scan system. This was the second major switch in voting equipment since the 2000 election. The volume of peripheral changes that have to be made to get staff up to speed on the new machines, get poll workers trained in using the equipment, providing materials to the voters to train them as well and then re-write all of your training manuals and procedures is massive. And, then we have the 2012 election with all of its issues. I mention all of this because out of everyone of these events, we accepted what was to be, learned from the changes and mistakes and were better for it. The report mentions several of the items already put into practice that we feel will alleviate any repeat of the problems we encountered in 2012. Changes in the election laws in 2013 that reversed what was previously enacted was a huge step forward. Our Board of County Commissioners worked very closely with us to fund the purchase of additional scanning equipment for future elections, purchasing electronic poll books to expedite the check-in process and finding a larger space for one of our branch offices used as an early voting site to accommodate the large number of voters in that area more efficiently. We have increased the number of voters using mail ballots through a strong campaign prior to our recent primary election which decreased the number of voters going to the polls on Election Day. Even though we do not think we will ever run into the same set of circumstances as we did in 2012, in the elections world we have learned to never say never. But, I sincerely believe that should similar instances repeat themselves, we will be more prepared and better able to handle the situation. If I can be of any additional assistance, please do not hesitate to contact me. Sincerely, Signed by: Sharon L. Harrington, CERA, MFCEP: Lee County Supervisor of Elections: PO Box 2545 (33902): 2480 Thompson Street (33901): Fort Myers, FL: 239-533-6301 (Phone): 239-533-6310 (Fax): Email: sharrington@leeelections.com: [End of section] Appendix V: GAO Contact and Acknowledgments: GAO Contact: Rebecca Gambler, (202) 512-8777 or gamblerr@gao.gov: Acknowledgments: In addition to the contact named above, Tom Jessor (Assistant Director), David Alexander (Assistant Director), Carl Barden (Assistant Director), Susan Czachor, Tony DeFrank, William Egar, Eric Hauswirth, Susan Hsu, Jeff Jensen, Elizabeth Kowalewski, Amanda Miller, Jan Montgomery, Rebecca Kuhlmann Taylor, Janet Temko-Blinder, Jeff Tessin, and Johanna Wong made significant contributions to this report. We gratefully acknowledge the substantial time and cooperation of the state and local election officials and researchers whom we interviewed. [End of section] Footnotes: [1] GAO analysis of data from the U.S. Census Bureau, Voting and Registration in the Election of November 2012--Detailed Tables, accessed July 1, 2013, [hyperlink, http://www.census.gov/hhes/www/socdemo/voting/publications/p20/2012/tabl es.html]. [2] See, for example, Charles Stewart III and Stephen Ansolabehere, "Executive Summary: Waiting in Line to Vote" (June 28, 2013), accessed July 9, 2013, [hyperlink, https://www.supportthevoter.gov/files/2013/08/Waiting-in-Line-to-Vote- White-Paper-Stewart-Ansolabehere.pdf]; and Douglas M. Spencer and Zachary S. Markovits, "Long Lines at Polling Stations? Observations from an Election Day Field Study," Election Law Journal, vol. 9, no. 1 (2010). [3] These include the National Voter Registration Act of 1993, the Help America Vote Act of 2002, and the Voting Rights Act of 1965, among others. [4] GAO, Elections: The Nation's Evolving Election System as Reflected in the November 2004 General Election, [hyperlink, http://www.gao.gov/products/GAO-06-450] (Washington, D.C.: June 6, 2006), and Elections: Perspectives on Activities and Challenges Across the Nation, [hyperlink, http://www.gao.gov/products/GAO-02-3] (Washington, D.C.: Oct.15, 2001). For our review of the 2004 general election, we surveyed a nationally representative random sample of 788 local election jurisdictions nationwide, stratified by population. For our review of the 2000 general election, we surveyed a random sample-- stratified by type of voting method--of (1) all county election jurisdictions, or their equivalents, in 39 states that delegate election responsibilities primarily to counties; (2) the largest minor civil divisions, such as towns and townships, in each county in the 9 states that delegate election responsibilities to these divisions; (3) the District of Columbia; and (4) Alaska. We excluded the state of Oregon from both surveys because it was a vote-by-mail state for these general elections. The survey for the 2000 general election was generalizable to this sample frame, which included 90 percent of the U.S. population. [5] While the presidential election process includes activities prior to Election Day--such as early voting--for the purposes of this review, we focus on voter wait times during in-person voting on Election Day. The results from our survey are generalizable to this population of jurisdictions nationwide, but our survey was not designed to have a sufficient sample to produce reliable estimates at the state level. We excluded jurisdictions with populations of 10,000 or fewer because, on the basis of our review of wait time research, jurisdictions of this size were unlikely to have experienced long voter wait times. We excluded the states of Oregon and Washington from our survey because, as of the November 2012 election, they were both vote-by-mail states. Because we followed a probability procedure based on random selections, our sample is only one of a large number of samples that we might have drawn. Since each sample could have provided different estimates, we express our confidence in the precision of our particular sample's results as a 95 percent confidence interval (e.g., from 5 to 15 percent). This is the interval that would contain the actual population value for 95 percent of the samples we could have drawn. [6] We conducted our survey from March 20, 2014, through June 15, 2014. To calculate our response rate, we used a standard definition, known as RR2, from the American Association for Public Opinion Research. See American Association for Public Opinion Research, 2011 Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys, 7th edition (2011). [7] We defined wait time as the time from when a voter entered the first line to when he or she began filling out a ballot. [8] The CCES is a survey of a nationally representative stratified sample of U.S. citizens aged 18 and over. The CCES has been conducted since 2006 to better understand congressional elections and representation using large-scale national surveys. The 2012 CCES surveyed 54,535 U.S. citizens aged 18 and over by Internet about their views and experiences before and after Election Day 2012. The project was the result of a collaborative effort of a consortium of research teams and organizations, and Stephen Ansolabehere of Harvard University was the principal investigator. [9] We also contacted election officials from the 3 remaining states, but they declined to be interviewed. [10] We interviewed officials from these jurisdictions between May and July 2014. [11] The Election Assistance Commission is an independent federal agency that was established in 2002 to help improve state and local administration of federal elections. [12] We identified relevant literature by searching social science, academic, and other databases for terms such as "voter wait times" and "election long lines," among others. [13] States primarily delegate election responsibilities to counties, but some delegate responsibilities to subcounty governmental units, such as townships or municipalities. [14] Pub. L. No. 107-252, 116 Stat. 1666 (2002) (codified as amended at 42 U.S.C. §§ 15301-545). [15] HAVA specifies that the EAC's four commissioners are to be nominated by the President on recommendations from Congress and confirmed by the U.S. Senate. [16] Pub. L. No. 103-31, 107 Stat. 77 (1993) (codified as amended at 42 U.S.C. §§ 1973gg-1973gg-10). [17] Pub. L. No. 89-110, 79 Stat. 437 (codified as amended at 42 U.S.C. §§ 1973 to 1973bb-1). [18] Collectively known as the language minority provisions of the Voting Rights Act, sections 203 and 4(f)(4) are to enable members of applicable language minority groups to participate effectively in the electoral process. 42 U.S.C. §§ 1973aa-1a, 1973b(f)(4). On the basis of 2010 Census results, 248 jurisdictions are covered under section 203 of the Voting Rights Act. The status of section 4(f)(4) is unclear as it relies on a coverage formula struck down by the Supreme Court in 2013. Shelby Co. v. Holder, 133 S. Ct. 2612 (2013). [19] States are required to perform regular maintenance of the voter list by comparing it against state records on felons and deaths and by matching voter registration application information on the voter list with information in the state motor vehicle agency's records and Social Security Administration records, as appropriate. [20] States can be divided into two groups according to how election responsibilities are delegated. The first group contains 41 states that delegate election responsibilities primarily to the county level, with a few of these states delegating election responsibilities to some cities, and 1 state that delegates these responsibilities to election regions. The District of Columbia is included in this group of states. The second group contains 9 states that delegate election responsibilities principally to subcounty governmental units. These local election jurisdictions vary widely in size and complexity, ranging from small New England townships to Los Angeles County, whose number of registered voters exceeds that of many states. [21] As of the 2012 general election, 35 states and the District of Columbia provided an opportunity for voters to cast a ballot prior to Election Day without providing an excuse, either by no-excuse absentee voting or by early voting, or both. [22] As of the 2012 general election, 27 states and the District of Columbia allowed for no-excuse absentee voting by mail, and 7 of these states and the District of Columbia allowed voters to apply for an absentee ballot on a permanent basis. [23] In some cases, multiple precincts may be combined in a single polling place.In addition, some states allow voters to cast their ballots at "vote centers," which are polling places at which any registered voter in the local election jurisdiction may vote on Election Day, regardless of the precinct in which the voter resides. According to the National Conference of State Legislatures, as of August 2014, 11 states have passed legislation that, for certain kinds of elections, has allowed jurisdictions to establish vote centers or has allowed vote center pilot projects in selected jurisdictions. [24] Jurisdictions call their poll workers by different titles, including clerks, wardens, election judges, inspectors, captains, and precinct officers and often have a chief poll worker for each polling place. On Election Day, poll workers set up and open polling places, which can include setting up the voting machines or voting booths, testing equipment, and posting required signs and voter education information. [25] Provisional ballots are those cast by voters at the polls whose eligibility to vote is unclear and to be determined later. HAVA requires states to provide a provisional ballot process for voters in certain circumstances. One such circumstance is when an individual asserts to be (1) registered in the jurisdiction for which he or she desires to vote and (2) eligible to vote in a federal election but (3) whose name does not appear on the official list of eligible voters for the polling place. Another is for first-time voters who register by mail but do not have required identification. [26] Some jurisdictions used hybrid systems that combine a DRE machine with an optical scanner; other jurisdictions used punch card systems, lever machines, or paper ballots that are manually cast and counted. [27] Some jurisdictions used DRE machines with a voter-verified paper audit trail, which prints out a paper record of the voters' ballot selections that is submitted for counting. [28] Exec. Order No. 13639, Establishment of the Presidential Commission on Election Administration, 78 Fed. Reg. 19979 (Mar. 28, 2013). See Presidential Commission on Election Administration, The American Voting Experience: Report and Recommendations of the Presidential Commission on Election Administration (January 2014). [29] According to the Presidential Commission report, the commission surveyed 7,779 local election officials from all 50 states, the District of Columbia, American Samoa, and Guam. The response rate for the survey was 41 percent. [30] The SPAE is conducted by Charles Stewart III, Professor of Political Science at the Massachusetts Institute of Technology. The total sample size for the 2012 SPAE survey was 10,200 people. [31] The total sample size for the 2012 CCES survey was 54,535. The survey consists of two phases in election years. In the preelection phase, respondents answer two-thirds of the questionnaire about general political attitudes, various demographic factors, assessment of roll call voting choices, and political information. In the postelection phase, respondents answer the other third of the questionnaire, mostly consisting of items related to the election that just occurred. In nonelection years, the survey consists of a single phase conducted in the early fall. [32] All estimates from the survey are subject to sampling error. The numbers in parentheses indicate results at the 95 percent confidence interval. [33] In our survey, we defined wait time as the time from when a voter entered the first line to when he or she began filling out a ballot. [34] See, for example, Schaefer Center for Public Policy, Voting and the Administration of Elections in Maryland (Baltimore, Maryland: January 2014), a report for the Maryland State Board of Elections. [35] See, for example, Charles Stewart III, "Waiting to Vote in 2012," Journal of Law and Politics, vol. 28, 439-463. [36] See Michael Herron and Daniel Smith, The Advancement Project, Congestion at the Polls: A Study of Florida Precincts in the 2012 General Election, (June 24, 2013), and Charles Stewart III, "Waiting to Vote in 2012." [37] See Schaefer Center for Public Policy, Voting and the Administration of Elections in Maryland. [38] Every jurisdiction in Maryland used electronic poll books and DRE voting systems for the 2012 election, making check-in and voting data available for each precinct in the state. According to researchers at the Schaefer Center, the simulation results identified similar counties where long wait times were more common on Election Day 2012 as those identified by the combined results of the CCES and SPAE surveys for the state. [39] See appendix II for wait time questions and response options. In our instructions to respondents, we noted that the time spent filling out or submitting a ballot may affect the wait time of later voters, but for the purposes of our survey, we wanted respondents to consider voter wait time to be only the time a voter waited prior to filling out a ballot. [40] As discussed earlier in this report, on the basis of our survey, we estimate that 16 percent (from 12 to 21 percent) of jurisdictions collected, received, or had available data that would allow them to calculate voter wait times that occurred at individual polling places. In addition, we estimate that some jurisdictions collected various types of wait time information that election officials and researchers have identified would be helpful in measuring wait times. [41] State election agencies, academic researchers, and other election experts have used different standards to define long voter wait times. For example, in its survey of local election jurisdictions, the Virginia State Board of Elections used a standard of 2 hours, or 120 minutes, or more to denote long wait times. The Presidential Commission did not define what constituted a long wait, but set a standard of 30 minutes as a target maximum wait time for voters. [42] In the remaining jurisdictions, officials believe that wait times of 10 minutes or less, more than 60 minutes, or more than 120 minutes are too long, or did not know what they considered to be too long. [43] For each response option that officials could select as a wait time they considered to be too long--ranging from 10 minutes or less to more than 120 minutes--the majority of jurisdictions said that no polling places had voter wait times that were too long on Election Day 2012. [44] Wait times may be based on officials' perspectives, data, or other information on wait times. [45] Because so few jurisdictions reported average voter wait times of more than 30 minutes at any time of Election Day 2012, we collapsed the response options of 21 to 30 minutes, 31 to 60 minutes, 61 to 120 minutes, and more than 120 minutes to create a single category of more than 20 minutes. Of the 32 jurisdictions that reported average wait times of more than 20 minutes during the hour before the polls closed, 15 of these reported average wait times of between 21 and 30 minutes and only 3 jurisdictions reported average wait times of 61 minutes or more. We reported the number of jurisdictions that reported these average wait times, rather than generalizable estimated percentages, because of the small number of jurisdictions involved. [46] Charles Stewart III conducted a similar analysis of CCES data and reported wait time estimates based on the combined responses of voters who voted early and also on Election Day 2012. Our analysis separated these combined responses to obtain estimates of voter wait times on Election Day. See Charles Stewart III, "Waiting to Vote in 2012." [47] Survey respondents were asked to estimate wait times within specified response categories. We replicated Charles Stewart III's approach in "Waiting to Vote in 2012" and estimated average wait times by first recoding the response categories to the midpoint of the category--for example, the "none at all" response was coded as 0 minutes, and the "1-10" response category was coded as 5 minutes. Respondents who waited more than an hour were asked to provide wait times in minutes. We excluded 12 states--Arkansas, Colorado, Hawaii, Mississippi, Montana, Nevada, New Hampshire, New Mexico, North Dakota, Rhode Island, Utah, and West Virginia--and the District of Columbia from our analysis because of the relative imprecision of their estimates (the estimated proportion of voters in these states who waited 10 minutes or less had a margin of error of greater than plus or minus 10 percentage points). [48] More specifically, we estimate that the percentage of voters who waited 61 minutes or more to vote were as follows: Florida--16 percent (from 12 to 21 percent), Maryland--12 percent (from 7 to 17 percent), and Virginia--12 percent (from 9 to 15 percent). [49] Charles Stewart III, "Waiting to Vote in 2012." Stewart analyzed data from the CCES and the SPAE, a nationally representative survey of 200 registered voters in each of the 50 states and the District of Columbia. The study reported wait time estimates based on the combined responses of voters who voted early and also on Election Day, but did not provide estimates broken out for each type of voting. [50] Schaefer Center for Public Policy, Voting and the Administration of Elections in Maryland. [51] All estimates from the survey are subject to sampling error. The numbers in parentheses indicate results at the 95 percent confidence interval. [52] See, for example, Stewart and Ansolabehere, "Executive Summary: Waiting in Line to Vote." This study found that among nonvoting respondents to the 2012 CCES, 0.8 percent stated that the main reason they did not vote was that "lines at the polls were too long." In addition, see Spencer and Markovits, "Long Lines at Polling Stations? Observations from an Election Day Field Study," 3. For this study, observers were assigned to monitor voter traffic at 30 polling places in three neighboring San Francisco Bay Area counties on Election Day for the 2008 California presidential primary election. The study found that 1.9 percent of the 11,858 voters in its sample stood in line for a period of time and then left without voting. [53] See appendix II for aggregated survey results related to these factors. [54] Totals do not sum to 100 because some jurisdictions may use both paper and electronic poll books. [55] Under federal law, a voter's registration remains active as long as the person resides at the address listed on his or her registration. Election officials can change a voter's registration status to inactive if the voter has moved and has not responded to an address confirmation request, among other circumstances. [56] Redistricting is the process of drawing new election district boundaries, which can affect precinct boundaries and polling place locations. Precinct consolidations generally involve combining two or more precincts into a single precinct and can result in changes to the locations of polling places. [57] As required by HAVA, a provisional ballot allows an individual to cast a ballot in an election with federal races despite questions regarding the individual's eligibility to vote. The provisional ballots are cast on ballots separate from the other ballots and examined for eligibility after the polls close. While provisional ballots are required by HAVA to be issued in certain circumstances, the methods of implementation are left to the discretion of the states and provisional ballots are counted in accordance with state law. [58] Some jurisdictions are required under section 203 of the Voting Rights Act to provide voting materials in specified minority languages in addition to English. This is determined by a prescribed statutory formula using the most recent Census data. 42 U.S.C. § 1973aa-1a. Certain jurisdictions are also required to provide bilingual voting materials under section 4(f)(4) of the act; however, that section relies on a coverage formula that was struck down by the Supreme Court in 2013. 42 U.S.C. § 1973b(f)(4); Shelby Co. v. Holder, 133 S. Ct. 2612 (2013). [59] Voting equipment refers to the voting systems employed by a jurisdiction for casting and counting votes, including electronic voting machines and paper balloting systems that use optical or digital scanners. [60] Federal laws that specifically address accessibility issues for voters with disabilities include the Americans with Disabilities Act of 1990 (Pub . L. No. 101-336, 104 Stat. 327, codified as amended at 42 U.S.C. §§ 12101 et seq.), HAVA, and the Voting Accessibility for the Elderly and Handicapped Act (Pub. L. No. 98-435, 98 Stat. 1678 (1984), codified at 42 U.S.C. §§ 1973ee to 1973ee-6), among others. For example among other requirements, when parking is available for voters, staff, and volunteers, accessible parking must be provided for people with disabilities. In addition, each polling place must have an accessible entrance connected to an accessible route. [61] See the results to question 9 of our survey in appendix II for more information on the types of data used by jurisdictions when allocating resources for Election Day 2012. [62] Jurisdiction officials said that state officials told them that as long as there was one working DRE voting machine in use at a precinct, that precinct could not utilize emergency paper ballots already supplied to the precinct that would need to be hand-counted, according to state law. Jurisdiction officials said that this law allowed paper ballots to be used in addition to DRE machines if the paper ballots would be counted by a machine, and not hand-counted. [63] A voter's registration status was marked as being inactive if prior to the election, the jurisdiction's voter registration office was unable to confirm the address of the voter by mail. According to the jurisdiction's postelection report, which was created by a bipartisan Election Process Task Force, there were a substantial number of inactive voters in 2012 because (1) the State Board of Elections did not comprehensively purge from voter rolls inactive voters who had become ineligible since the last presidential election, and (2) many voters had their registration status changed to inactive as a result of returned mailings from materials sent out as part of the State Board of Election's voter education campaign to inform voters of precinct changes associated with redistricting. [64] There were 251 forms filled out in 2008 and 3,100 in 2012, according to jurisdiction data. The jurisdiction's postelection report noted that the increase in the number of address confirmations required in 2012 was primarily because of the large number of inactive voters, although the jurisdiction's General Registrar noted that voters were required to fill out Affirmation of Eligibility forms for other reasons also. [65] The postelection report was created by a bipartisan Election Process Task Force that included a number of private citizens representing a wide range of experience and participation in county affairs. This task force was assisted by county staff. [66] In response to long voter wait times experienced during the 2008 presidential election, the state issued guidance on calculating the number of voting booths a polling place may need. The guidance instructs jurisdictions to take into consideration anticipated voter turnout based on 2008 numbers, the length of the ballot, and the number of voters a voting booth can process per hour. [67] According to jurisdiction officials, when they became aware of the length of the ballot for the 2012 presidential election, testing was conducted on different constituencies, including the elderly and groups with varying education levels, to determine the length of time it might take voters in the jurisdiction to mark their ballots. [68] Jurisdiction officials stated that in some precincts, they set up one voting station for every 40 to 60 expected voters. This exceeded both state law requirements, which mandated one voting station for every 300 registered voters, and state guidance, which recommended one voting booth for every 80 to 100 voters. [69] According to jurisdiction officials, the state redistricted in response to the 2010 decennial Census results and this led to the jurisdiction redrawing precinct boundaries, which affected polling place locations. [70] Some states allow for Election Day registration, which permits any qualified resident of the state to register and vote on Election Day. [71] The National Voter Registration Act of 1993 requires that motor vehicle agencies and agencies that provide public assistance offer eligible individuals the opportunity to register to vote, among other things. [72] State law generally requires board members of fire control, mosquito control, and community development districts to be elected in general elections. The jurisdiction's Supervisor of Elections stated that community development districts are similar to residential housing associations and that these districts accounted for 44 of the 54 total races on the jurisdiction's ballots. [73] This jurisdiction was required under section 203 of the Voting Rights Act to provide voting materials in specified minority languages in addition to English. See 42 U.S.C. § 1973aa-1a. [74] For the 2008 election, state law required jurisdictions to offer between 12 and 14 days of early voting and at least 8 hours on each weekday during the authorized period. For the 2012 election, state law required 8 days of early voting and between 6 and 12 hours each day. [75] For both the 2008 and 2012 elections, state law allowed jurisdictions to have early voting at city halls, public libraries, and main or permanent branch offices of the supervisors of elections. [76] According to election officials, the state legislature redistricted in response to 2010 decennial Census results and this led to the jurisdiction redrawing precinct boundaries, which affected polling place locations. Precinct consolidations generally involved combining 2 or more precincts into a single precinct and resulted in changes to the locations of polling places. [77] In 2013, a state law was enacted that limited ballot summaries for amendments proposed by a joint resolution to 75 words for the first measure on the ballot and no longer allows the full text of the amendment to be used instead of a ballot summary. The word limit does not apply to ballot summary revisions to correct deficiencies identified by a court. In the 2012 election, ballot summaries for amendments or other public measures were subject to a 75-word limit, but this did not apply to those proposed by a joint resolution. [78] In 2013, a state law was enacted that requires jurisdictions to offer between 8 and 14 days of early voting in an election that contains a state or federal race and between 8 and 12 hours per voting day. In addition, the law added fairgrounds, civic centers, courthouses, stadiums, convention centers, and other types of locations as permissible early voting sites. [79] According to officials, provisional ballots have increased in part because many voters who signed up for mail-in voting-- particularly permanent absentee voters--did not realize that they had done so and came to the polls to vote in person. [80] Federal, state, county, and applicable local elected offices and ballot questions are counted for the provisional ballots of individuals who vote in a precinct other than the one to which they are assigned. If the ballot cast by the voter contains the same offices and questions on which the voter would have been entitled to vote in his or her assigned precinct, the votes for the entire ballot are counted. If the ballot cast by the voter contains offices or questions on which the voter would not have been entitled to vote in his or her assigned precinct, only the votes for which the voter was entitled to vote in his or her assigned precinct are counted. Jurisdiction officials noted that eligibility for votes to be counted is determined after the polls close. This state's policy is in contrast to the policies of states where provisional ballots must be cast in the precinct in which voters are registered for ballots to be eligible to be counted. [81] State law requires that when precinct boundaries are established or changed, the number of voters in the precinct must not exceed 1,000 on the 88th day prior to the day of election, unless otherwise provided by law. [82] State law requires the Secretary of State to mail sample ballots to registered voters between 21 and 40 days before the election. [83] The jurisdiction conducts an annual survey of polling place inspectors. According to its 2012 report on survey results, nearly 95 percent of inspectors said that training prepared them for Election Day and about 91 percent of inspectors rated their fellow poll workers positively as either excellent or very good. [84] The results from our survey are generalizable to this population of jurisdictions. [85] The CCES is a survey of a nationally representative stratified sample of U.S. citizens aged 18 and over. The CCES has been conducted since 2006 to better understand congressional elections and representation using large-scale national surveys. The 2012 CCES surveyed 54,535 U.S. citizens aged 18 and over by Internet about their views and experiences before and after Election Day 2012. The project was the result of a collaborative effort of a consortium of research teams and organizations, and Stephen Ansolabehere of Harvard University was the principal investigator. [86] We also contacted election officials from the 3 remaining states, but they declined to be interviewed. [87] We identified relevant literature using search terms such as "voter wait times" and "election long lines," among others, in various databases, including Academic OneFile, Dissertation Abstracts, JSTOR, PolicyFile, and Social SciSearch. [88] As with all surveys that rely on self-reported information, estimates may be imprecise or responses may be subject to recall error, if based on recollections. [89] To calculate our response rate, we used a standard definition, known as RR2, from the American Association for Public Opinion Research. See American Association for Public Opinion Research, 2011 Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys, 7th edition (2011). [90] The county equivalents for Alaska were assigned to their respective election regions. [91] We did not use numbers of registered voters to define the strata because Census data on registered voters were not available at the county and MCD levels nationwide. We also did not use numbers of eligible voters 18 years and over to define the strata because Census data allowing us to exclude noncitizens and felons from the 18 years and over population were also not available at the county and MCD levels nationwide. Noncitizens are not eligible to vote, and voting eligibility for citizens convicted of a felony varies among states. [92] We applied weighting-class adjustments for nonresponse by multiplying the base sampling weights with the inverse of the stratum response rates. [93] We explored analyzing our survey to determine how voter wait times were associated with selected demographic characteristics and election administration policies and practices. Limited variation among reported wait times, with most jurisdictions reporting no or minimal wait times, made this analysis infeasible because we could not estimate the associations with enough precision. For example, on the basis of officials' survey responses, we estimate that 79 percent (from 75 to 84 percent) of jurisdictions had no polling places and 9 percent (from 6 to 13 percent) had only a few polling places with wait times of greater than 60 minutes on Election Day 2012. [94] See, for example, GAO, Elections: The Nation's Evolving Election System as Reflected in the November 2004 General Election, [hyperlink, http://www.gao.gov/products/GAO-06-450] (Washington, D.C.: June 6, 2006), and Elections: Perspectives on Activities and Challenges Across the Nation, [hyperlink, http://www.gao.gov/products/GAO-02-3] (Washington, D.C.: Oct. 15, 2001). [95] The results of GAO's 2014 survey of local election jurisdictions are generalizable to all jurisdictions nationwide, excluding those with populations of 10,000 or fewer and jurisdictions in Oregon and Washington. The survey was not designed to have a sufficient sample to produce reliable estimates at the state level. [96] Charles Stewart III conducted a similar analysis of CCES data and reported wait time estimates based on the combined responses of voters who voted early and also on Election Day 2012. Our analysis separated these combined responses to obtain estimates of voter wait times on Election Day. See Charles Stewart III, "Waiting to Vote in 2012," Journal of Law and Politics, vol. 28, 439-463. [97] We also contacted election officials from the 3 remaining states, but they declined to be interviewed. [98] The Election Assistance Commission is an independent federal agency that was established by the Help America Vote Act of 2002 to help improve state and local administration of federal elections. [99] The 2012 CCES survey was conducted in two phases. The preelection phase was conducted during October 2012, and gauged issue preferences, knowledge of the candidates, and voter intentions. The postelection phase was administered in November 2012, following Election Day (November 6, 2012), and asked, among other things, whether or not respondents voted, reasons why (if they did not vote), and questions about their voting experience. Responses to the postelection survey were based on voter perspectives of their experiences, which could be subject to recall error or influenced by media coverage. [100] Stephen Ansolabehere, Cooperative Congressional Election Study, 2012, Common Content: Release 1 (Cambridge, Massachusetts, Harvard University: April 15, 2013). [End of section] GAO's Mission: The Government Accountability Office, the audit, evaluation, and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO's website [hyperlink, http://www.gao.gov]. Each weekday afternoon, GAO posts on its website newly released reports, testimony, and correspondence. To have GAO e-mail you a list of newly posted products, go to [hyperlink, http://www.gao.gov] and select "E-mail Updates." Order by Phone: The price of each GAO publication reflects GAO's actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO's website, [hyperlink, http://www.gao.gov/ordering.htm]. Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537. Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information. Connect with GAO: Connect with GAO on facebook, flickr, twitter, and YouTube. Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts. Visit GAO on the web at [hyperlink, http://www.gao.gov]. To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; E-mail: fraudnet@gao.gov; Automated answering system: (800) 424-5454 or (202) 512-7470. Congressional Relations: Katherine Siggerud, Managing Director, siggerudk@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, DC 20548. Public Affairs: Chuck Young, Managing Director, youngc1@gao.gov: (202) 512-4800: U.S. Government Accountability Office: 441 G Street NW, Room 7149: Washington, DC 20548. [End of document]