This is the accessible text file for GAO report number GAO-12-69 entitled 'Elections: Views on Implementing Federal Elections on a Weekend' which was released on January 12, 2012. 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 Committees: January 2012: Elections: Views on Implementing Federal Elections on a Weekend: GAO-12-69: GAO Highlights: Highlights of GAO-12-69, a report to congressional committees. Why GAO Did This Study: Many U.S. citizens who are eligible to vote in federal elections do not do so. For instance, in the 2008 general election, about 62 percent of eligible citizens voted. To increase voter turnout by enhancing convenience, some states have implemented alternative voting methods, such as in-person early voting-—casting a ballot in person prior to Election Day without providing a reason—-and no-excuse absentee voting-—casting an absentee ballot, usually by mail, without providing a reason. In general, since 1845, federal law has required that federal elections be held on Tuesday. The committees on appropriations directed GAO to study and report on costs and benefits of implementing H.R. 254—-the Weekend Voting Act—- including issues associated with conducting a weekend election. Specifically, this report addresses: (1) alternatives to voting on Tuesday that states provided for the November 2010 general election, (2) how election officials anticipate election administration and costs would be affected if the day for federal elections were moved to a weekend, and (3) what research and available data suggest about the potential effect of a weekend election on voter turnout. GAO reviewed H.R. 254 and analyzed state statutes and early voting turnout in the 2010 Maryland elections, which had early voting over weekdays and weekends. GAO interviewed election officials in nine states, the District of Columbia (District), and 17 local jurisdictions that were selected on the basis of geographic dispersion and experience with weekend voting, among other things. Though not generalizable, the interviews provide insights. What GAO Found: For the 2010 general election, 35 states and the District provided voters at least one alternative to casting their ballot on Election Day through in-person early voting, no-excuse absentee voting, or voting by mail. Specifically, 33 states and the District provided in- person early voting, 29 states and the District provided no-excuse absentee voting, and 2 states provided voting by mail to all or most voters. Of the 9 states and the District where GAO conducted interviews, all but 2 states provided voters the option of in-person early voting in the 2010 general election, and 5 states and the District offered both early voting and no-excuse absentee voting. Implementation and characteristics of in-person early voting varied among the 7 states and, in some cases, among the jurisdictions within a state. For example, 5 states and the District required local jurisdictions to include at least one Saturday, and 2 states allowed for some jurisdiction discretion to include weekend days. State and local election officials GAO interviewed identified challenges they would anticipate facing in planning and conducting Election Day activities on weekends-—specifically, finding poll workers and polling places, and securing ballots and voting equipment—- and expected cost increases. Officials in all 17 jurisdictions and the District we contacted said they expected the number of poll workers needed for a 2-day weekend election would increase. Further, officials in 13 jurisdictions said that some poll workers would be less willing to work on the weekend because of other priorities, such as family obligations or attending religious services. Officials in 14 of the 17 jurisdictions and the District expected that at least some of the polling places they used in past elections-—such as churches—-would not be available for a weekend election, and anticipated difficulty finding replacements. Officials in all 9 states, the District, and 15 of the 17 local jurisdictions said ensuring the security of ballots and voting equipment over the Saturday night of a weekend election would be both challenging and expensive. Officials in 5 of the 7 states and the District that conducted early voting and provided security over multiple days explained that the level of planning needed for overnight security for a weekend election would far surpass that of early voting due to the greater number and variety of Election Day polling places. For example, officials in one state said that for the 2010 general election, the state had fewer than 300 early voting sites-—which were selected to ensure security-—compared to more than 2,750 polling places on Election Day, which are generally selected based on availability and proximity to voters. In addition, officials in all 9 states, the District, and 15 of the 17 local jurisdictions said they expected overnight security costs to increase. Weekend elections have not been studied, but studies of other voting alternatives determined that voter turnout is not strongly affected by them. Since nationwide federal elections have never been held on a weekend, it is difficult to draw valid conclusions about how moving federal elections to a weekend would affect voter turnout. GAO’s review of 24 studies found that, with the exception of vote by mail, each of the alternative voting methods was estimated to change turnout by no more than 4 percentage points. GAO’s analysis of early voter turnout data in Maryland found that 1.5 percent of voters we analyzed cast ballots on the weekend during the 2010 general election. View [hyperlink, http://www.gao.gov/products/GAO-12-69]. For more information, contact William O. Jenkins, Jr., at (202) 512-8777 or jenkinswo@gao.gov. [End of section] Contents: Letter: Background: Most States Provided Early or No-Excuse Absentee Voting as Alternatives to Voting on Tuesday in the 2010 General Election: Most Election Officials We Interviewed Expect Greater Difficulty and Costs Associated with a Weekend Election: Weekend Elections Have Not Been Studied, but Studies of Other Voting Alternatives Suggest That Voter Turnout May Not Be Strongly Affected: Appendix I: Selected Characteristics of States and Local Election Jurisdictions We Contacted: Appendix II: Analysis of Weekend Voting in the 2010 Maryland State Elections: Appendix III: Alternative Voting Methods Provided in 50 States and the District for the 2004 and 2010 November General Elections: Appendix IV: Selected Details of Early and No-Excuse Absentee Voting for the 2010 General Election in States We Contacted: Appendix V: Selected Details of Early Voting for the 2010 General Election in Local Jurisdictions We Contacted: Appendix VI: GAO Contact and Staff Acknowledgments: Bibliography: Related GAO Products: Tables: Table 1: Estimated Impacts of Alternative Voting Methods on Voter Turnout: Table 2: Turnout and Use of Saturday Poll Hours in the 2010 Maryland General Election: Table 3: Selected Information about the States and the District We Contacted: Table 4: Selected Information about the Local Election Jurisdictions and the District We Contacted: Table 5: Turnout and Use of Weekend Poll Hours in 2010 Maryland General Election, by Demographic Group: Table 6: Selected Details of Early and No-Excuse Absentee Voting Implementation for the November 2010 General Election in States We Contacted: Table 7: Selected Details of Early Voting Implementation for the November 2010 General Election in Local Jurisdictions We Contacted: Figures: Figure 1: The Election Process Involves the Integration of People, Process, and Technology: Figure 2: Thirty-Five States and the District Provided at Least One Alternative to Voting on Tuesday in the November 2010 General Election: Figure 3: Early Voting in Maryland's November 2010 General Election: Figure 4: In-person Early Voting and No-excuse Absentee Voting in 50 States and the District for the 2004 and 2010 General Elections: Abbreviations: District: District of Columbia: DRE: direct recording electronic: EAC: Election Assistance Commission: EVIC: Early Voting Information Center: HAVA: Help America Vote Act: NASS: National Association of Secretaries of State: NCSL: National Conference of State Legislatures: United States Government Accountability Office: Washington, DC 20548: January 12, 2012: The Honorable Richard Durbin: Chairman: The Honorable Jerry Moran: Ranking Member: Subcommittee on Financial Services and General Government: Committee on Appropriations: United States Senate: The Honorable Jo Ann Emerson: Chairwoman: The Honorable José Serrano: Ranking Member: Subcommittee on Financial Services and General Government: Committee on Appropriations: House of Representatives: Voting is fundamental to our democracy, yet many U.S. citizens who are eligible to vote do not take advantage of their constitutional right to vote in federal elections. For instance, about 62 percent of eligible citizens voted in the November 2008 general election and about 42 percent voted in the November 2010 general election.[Footnote 1] Policymakers, legislators, researchers, and advocacy organizations have explored ways to make voting more convenient with the goal of increasing voter turnout. Some states have implemented alternative voting methods aimed at increasing convenience by offering citizens an alternative to voting on Election Day. These include in-person early voting--that is, casting a ballot in person prior to Election Day without providing a reason, and no-excuse absentee voting--that is, casting an absentee ballot, usually by mail, without providing a reason. Since 1845, federal law has required that federal Election Day for the offices of President and Vice President be held on Tuesday.[Footnote 2] Subsequent legislation extended the Tuesday Election Day to House and Senate elections.[Footnote 3] Thus, while states or local jurisdictions may have held nonfederal elections on a weekend or conducted early voting during a period that included Saturdays or Sundays, nationwide general elections for federal offices have been held on Tuesdays. Various bills have been introduced in Congress since 1995 to increase voter convenience by changing the day of presidential and congressional elections from Tuesday to the weekend.[Footnote 4] Specifically, legislation has been introduced to move the day for regularly scheduled federal elections from the Tuesday after the first Monday in November to the Saturday and Sunday after the first Friday in November, and to synchronize the opening and closing times of polling places within the continental United States for federal elections. As of December 2011, the most recent bill, introduced in January 2009, was H.R. 254--the Weekend Voting Act. The committees on appropriations directed us to study and report on costs and benefits of implementing H.R. 254, including the anticipated costs to state and federal election administration officials, the effects on polling places, and the estimated increase in voter turnout.[Footnote 5] Since 2001, we have issued a series of reports covering aspects of the election process primarily with respect to federal elections. For example, after the November 2000 and 2004 elections, we issued comprehensive reports on processes involved in and challenges associated with administering those elections.[Footnote 6] This report focuses on issues related to implementing weekend elections as described in H.R. 254. Specifically, it addresses the following questions: * What alternatives to voting on Tuesday did states provide for the November 2010 general election? * How, if at all, do election officials anticipate election administration and associated costs would be affected if the day for federal elections were moved to a weekend? * What do research and available data suggest about the potential effect of a weekend election on voter turnout? To address our objectives, we reviewed H.R. 254, the U.S. Election Assistance Commission's (EAC) study of Alternative Voting Methods and 2010 Statutory Overview report,[Footnote 7] the Center for Democracy and Election Management's Election Administration Profiles of All Fifty States,[Footnote 8] and states' and local jurisdictions' election websites. We also conducted interviews with election officials in a nonprobability sample of nine states and the District of Columbia (District), and a nonprobability sample of 17 local jurisdictions within those states, about if and how they implemented alternative voting methods and their views on how election administration and voter turnout would likely be affected in their state or jurisdiction if the day for regularly scheduled federal elections were moved to a weekend.[Footnote 9] To obtain a range of perspectives, we selected states that varied according to, among other things, geographic region, experience with voting on weekends (such as with early voting on Saturdays or Sundays) and alternative voting methods in federal elections, and the level of local government responsible for administering elections (e.g., county or township). To obtain different perspectives at the local level, we selected 17 jurisdictions to reflect variation in factors including demographics (e.g., population and median household income), experience with weekend voting, and other characteristics such as having a large military or student population. Appendix I provides information about the states and 17 local jurisdictions we selected. The results of our interviews cannot be generalized to all states and local election jurisdictions; however, they provide important information about how alternative voting methods were implemented in different states and local jurisdictions. They also provide perspectives on how, if at all, elements of elections administration (e.g., voter registration, poll worker recruitment, security of voting equipment, and costs) or voter turnout might be affected if the day for regularly scheduled federal elections were on a weekend. Moreover, to address objective 1, we analyzed state statutes regarding three such methods--in-person early voting, no-excuse absentee voting, and voting by mail.[Footnote 10] Specifically, we assessed which of the 50 states and the District provided these methods for the November 2010 general election. We used EAC's 2010 Statutory Overview as a source for the state statute references, as well as our own research, and also reviewed information provided about these three methods from organizations that have compiled election information--the EAC, Early Voting Information Center (EVIC), National Association of Secretaries of State (NASS), and National Conference of State Legislatures (NCSL). If we needed clarification regarding a state statute, we called the relevant Secretary of State's or election office. To see whether and how the number of states that provided each alternative voting method changed between November 2004 and 2010, we compared the results of our analysis with information we previously reported on the 2004 general election.[Footnote 11] We supplemented information on the 50 states with additional details, such as whether the early voting period included weekend hours, which we obtained from election Web sites of the selected states and local jurisdictions as well as interviews with and documents provided by those election officials. To address objective 2, we interviewed state and local election officials. When available, we obtained information on costs, such as poll worker costs, that jurisdictions incurred for the 2010 general election, as well as officials' estimates of what their costs might be if the day for federal elections were moved to a weekend. However, because no state or jurisdiction had experience with conducting a federal general election on a weekend, the estimates we obtained are uncertain. To obtain additional views, we attended conferences and meetings held by the National Association of State Election Directors, Joint Election Officials Liaison Committee, and National Association of Counties, and solicited views of attending members. We also interviewed representatives of these organizations as well as elections experts and researchers from the EAC, Election Center, EVIC, Pew Center on the States, NCSL, League of Women Voters, Why Tuesday?, [Footnote 12] and the University of Maryland. To address objective 3, we (1) identified methodological challenges one would encounter when determining the effect; (2) reviewed what researchers have concluded about how alternative voting methods have influenced voter turnout; and (3) analyzed voter turnout in Maryland, which conducted early voting on weekdays and weekends in the November 2010 elections. To describe challenges, we identified jurisdictions in the United States that have experimented with weekend voting and assessed the availability of their voter turnout and demographic data. However, because federal elections nationwide have never been held on a weekend, we could not use historical data to directly estimate how changing the day of federal elections would affect voter turnout. To review research on how voting methods affect turnout, we searched databases of political science journals and reports of foundations and think tanks with expertise in election reform published since 1985. [Footnote 13] We identified 24 studies that estimated how alternative voting methods have affected turnout in state and local elections in the United States from 1972 through 2008 and collected the estimates from each study.[Footnote 14] To analyze voter turnout, we estimated how turnout varied in Maryland between weekday and weekend early voting periods and whether certain demographic groups of voters, such as those who voted in previous elections, were more likely to vote on the weekend. We selected Maryland for our analysis from among the nine states where we conducted interviews because they maintained reliable and detailed electronic data on voter turnout and voter registration. We interviewed state election officials responsible for maintaining the data and conducted electronic tests for logic and accuracy, and we determined that the data were sufficiently reliable for our purposes. However, the experiences of Maryland do not predict how turnout would change if federal elections were held on weekends nationwide. Further, our analysis was unable to account for several factors, including differences in polling hours across jurisdictions and voter demographics, such as education, not measured by state voter databases. Appendix II describes our data analysis. We conducted this performance audit from December 2010 to January 2012 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: Election Authority: The basic goal of the elections system in the United States is that all eligible voters have the opportunity to cast their vote and have their valid ballot counted accurately. Election authority is shared by federal, state, and local officials, and the election system is highly decentralized. States are responsible for the administration of their own elections as well as federal elections, and states regulate various aspects of elections including registration procedures, absentee voting requirements, alternative voting methods, establishment of polling places, provision of Election Day workers, testing and certification of voting equipment, and counting and certification of the vote.[Footnote 15] As the U.S. election system is highly decentralized, primary responsibility for managing, planning, and conducting elections resides locally with about 10,500 local election jurisdictions nationwide. In most states, election responsibility resides at the county level, although some states have delegated election responsibility to subcounty governmental units, such as cities, villages, and townships.[Footnote 16] Local election jurisdictions vary widely in size and complexity, ranging from small New England townships to Los Angeles County, where the number of registered voters exceeds that of 42 states. Some states have mandated statewide election administration guidelines and procedures that foster uniformity in the way 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, it is local jurisdictions that pay for elections. According to the Executive Director of the EAC, costs are not tracked in uniform ways because of the decentralized nature of elections and the variation in state and jurisdiction size and funding structures. However, Congress has authority to affect the administration of elections in certain ways. Congress' authority to regulate elections derives from various constitutional sources, depending upon the type of election.[Footnote 17] Federal legislation has been enacted in major functional areas of the voting process, such as voter registration, absentee voting requirements, accessibility provisions for the elderly and voters with disabilities, and prohibitions against discriminatory voting practices. With regard to the administration of federal elections, Congress has constitutional authority over both presidential and congressional elections, including the timing of federal elections.[Footnote 18] Under federal statute, the Tuesday after the first Monday in November in an even-numbered year is established as the day for federal congressional elections.[Footnote 19] Federal statute also sets this same day for the selection of presidential electors--the Tuesday after the first Monday in November in every 4th year succeeding every election of a President and Vice President.[Footnote 20] In general, these are the federal statutes that the previously pending weekend voting bills would have amended to move the November Tuesday federal Election Day to Saturday and Sunday. Such a change in federal law would, in effect, likely require states to change their laws and regulations governing the implementation of federal elections to mirror the day(s) established in federal law. Current federal law does not dictate the hours that polling places are required to be open on Election Day. The timing of state and local elections is not mandated by the federal election calendar. Nevertheless, many state and local government officials are also elected on federal Election Day as a matter of convenience and to save costs. According to the EAC, some states and local jurisdictions have held nonfederal elections or primaries on Saturdays, believing that it might be more convenient for voters and, in turn, might increase voter turnout.[Footnote 21] For example, in Louisiana, all nonfederal elections take place on Saturdays and, in Texas, some nonfederal elections such as general elections for cities and schools take place on Saturdays. From 1978 through 2006, Delaware held local elections, including primaries, on Saturdays. It held its first Saturday presidential primary in 1996. However, according to the EAC, because the Jewish Sabbath is on Saturday and, additionally, the state's 2002 primary fell on the Jewish New Year, Delaware moved the presidential primary to Tuesday in 2004 and the state primary to Tuesday in 2006. Election System Elements: The U.S. election system is based on a complex interaction of people (voters, election officials, and poll workers), process, and technology that must work effectively together to achieve a successful election, as shown in figure 1. Figure 1: The Election Process Involves the Integration of People, Process, and Technology: [Refer PDF for image: illustration] People and Technology are integrated into each aspect of Process: Process: Voter registration; Absentee and early voting; Planning and conducting Election Day activities; Voting with provisional ballots; Vote counting and certification. Source: GAO. [End of figure] The election process is dependent on the citizens who cast ballots; however, election officials and poll workers are also essential to making the system work.[Footnote 22] State and local election officials are either elected or appointed and are responsible for carrying out federal and state election requirements. This can be a year-round effort. Among other things, election officials register eligible voters and maintain voter registration lists; design ballots; educate voters on how to vote; 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. However, elections also depend on an army of poll workers--about 2 million for a federal election--who are willing to staff the polls on Election Day. Some poll workers are elected, some are appointed by political parties, and some are volunteers. Compensation varies by the level of responsibility of the poll worker and the state or jurisdiction in which they work. As we reported in 2006, increasingly, poll workers are needed with different skills, such as computer or technical skills, and across the country jurisdictions have faced challenges finding poll workers.[Footnote 23] Voting methods and related technology also play a critical part in the success of an election. Voting methods are tools for accommodating the millions of voters in our nation's approximately 10,500 local election jurisdictions. Since the 1980s, ballots in the United States, to varying degrees, have been cast and counted using five methods: paper ballots, lever machines, punch cards, optical scan,[Footnote 24] and direct recording electronic (DRE) machines.[Footnote 25] Four of these methods involve technology; only the paper ballot system does not. For example, many DREs use computers to present the ballot to the voter, and optical scan and DRE systems depend on computers to tally votes. The way voting systems are designed, developed, tested, installed, and operated can lead to a variety of situations where misunderstanding, confusion, error, or deliberate actions by voters or election workers can, in turn, affect the equipment's performance in terms of accuracy, ease of use, security, reliability, and efficiency. Each of the 50 states and the District has its own election system with a somewhat distinct approach. While election systems vary from one local jurisdiction to another, all involve people, process, and technology, and most have the following elements: * Voter registration. Voter registration is not a federal requirement. However, except for North Dakota, all states and the District generally require citizens to register before voting. The deadline for registering and what is required to register varies. At a minimum, state eligibility provisions typically require a person to be a U.S. citizen, at least 18 years of age, and a resident of the state, with some states requiring a minimum residency period. Citizens apply to register to vote in various ways, such as at motor vehicle agencies, by mail, or at local voter registrar offices. Some states allow citizens to register at a polling place on Election Day. Election officials process registration applications and compile and maintain the list of registered voters to be used throughout the administration of an election. * Absentee and early voting. Absentee voting is a process that allows citizens the opportunity to vote when they are unable to vote at their precinct on Election Day and is generally conducted by mail. All states and the District have provisions allowing voters to cast their ballot before Election Day by voting absentee with variations on who may vote absentee, whether the voter needs an excuse, and the time frames for applying and submitting absentee ballots.[Footnote 26] In addition, some states also allow in-person early voting, as discussed later in the report. In general, early voting allows voters from any precinct in the jurisdiction to cast their vote in person without an excuse before Election Day either at one specific location or at one of several locations. Early voting locations have a registration list for the jurisdiction and ballots specific to each precinct. The voter is provided with and casts a ballot designed for his or her assigned precinct. As with absentee voting, the specific circumstances for in- person early voting--such as the dates, times, and locations--are based on state and local requirements. * Planning and conducting Election Day activities. Election officials perform a range of activities in preparation for and on Election Day itself. Prior to an election, officials recruit and train poll workers to have the skills needed to perform their Election Day duties, such as opening and closing the polls and operating polling place equipment. Where needed and required, election officials must also recruit poll workers who speak languages other than English.[Footnote 27] Officials also locate polling places that are to meet basic standards for accessibility and have an infrastructure to support voting machines as well as voter and poll worker needs.[Footnote 28] They design and produce ballots to meet state requirements and voter language needs, and that identify all election races, candidates, and issues on which voters in each precinct in their jurisdiction will vote. Election officials seek to educate voters on topics such as what the ballot looks like, how to use a voting machine, and where their particular polling place is located. Finally, election officials seek to ensure that voting equipment, ballots, and supplies are delivered to polling places. On Election Day, poll workers set up and open the polling places. This can include setting up the voting machines or voting booths, testing equipment, posting required signs and voter education information, and completing paperwork such as confirming that the ballot is correct for the precinct. Before a voter receives a ballot or is directed to a voting machine, poll workers typically are to verify his or her eligibility. * Provisional voting. Federal law requires that an individual asserting to be registered in the jurisdiction for which he or she desires to vote and is eligible to vote in a federal election--but whose name does not appear on the official list of eligible voters for the polling place--be provided a provisional ballot. In addition, provisional ballots are to be provided in elections for federal office to individuals whom an election official asserts to be ineligible to vote, and for court-ordered voting in a federal election after the polls have closed. 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. * Vote counting and certification. Following the close of the polls, election officials and poll workers complete steps to count the votes and determine the outcome of the election. Equipment and ballots are to be secured, and votes are to be tallied or transferred to a central location for counting. The processes used to count or to recount election votes vary with the type of voting equipment used in a jurisdiction, state statutes, and local jurisdiction policies. Votes from Election Day, absentee ballots, early votes (where applicable), and provisional ballots are to be counted and consolidated for each race to determine the outcome. While preliminary results are available usually by the evening of Election Day, the certified results are generally not available until days later. Most States Provided Early or No-Excuse Absentee Voting as Alternatives to Voting on Tuesday in the 2010 General Election: For the November 2010 general election, 35 states and the District provided voters at least one alternative to casting their ballot on Election Day through in-person early voting, no-excuse absentee voting, or voting by mail. As shown in figure 2, 33 states and the District provided in-person early voting, 29 states and the District provided no-excuse absentee voting, and 2 states provided voting by mail to all or most voters.[Footnote 29] Figure 2: Thirty-Five States and the District Provided at Least One Alternative to Voting on Tuesday in the November 2010 General Election: [Refer to PDF for image: illustrated U.S. map] In-person early voting: Arkansas: Indiana: Louisiana: Tennessee: Texas: West Virginia: In-person early voting and no-excuse absentee voting: Alaska: Arizona: California: Colorado: District of Columbia: Florida: Georgia: Hawaii: Idaho: Illinois: Iowa: Kansas: Maine: Maryland: Montana: Nebraska: Nevada: New Jersey: New Mexico: North Carolina: North Dakota: Ohio: Oklahoma: South Dakota: Utah: Vermont: Wisconsin: Wyoming: Vote by mail: Oregon: Washington: Source: GAO analysis of alternative voting methods for the November 2010 general election (data); MapQuest (map). Note: The November 2010 election was conducted entirely by mail in Oregon and in 38 of 39 counties in Washington. In the remaining Washington county (Pierce County), voters were allowed to cast ballots in person at polling places on Election Day. Washington and Oregon are shown in figure 2 as vote by mail states but also had provisions that met our definition of no-excuse absentee voting and are included in our no-excuse absentee voting numbers. Both states also allowed voters to cast their vote-by-mail ballot in person at designated ballot deposit sites prior to Election Day without providing an excuse. [End of figure] In addition, eight of the states and the District with no-excuse absentee voting permitted registered voters to apply for an absentee ballot on a permanent basis so those voters automatically receive an absentee ballot in the mail prior to every election without providing an excuse or reason for voting absentee.[Footnote 30] Furthermore, the number of states providing these alternatives has increased in recent elections. We previously reported that for the 2004 general election, 24 states and the District required or allowed in-person early voting, 21 states required or allowed no-excuse absentee voting, and 1 state--Oregon--required all voters to vote by mail.[Footnote 31] Appendix III compares the alternative voting methods for the 2004 and 2010 general elections, by state. Of the nine states and the District where we conducted interviews, all but two states provided voters the option of in-person early voting in the November 2010 general election. Five of the seven states and the District offered both early voting and no-excuse absentee voting. Appendix IV provides additional details of how these seven states and the District implemented these two alternative voting methods for the 2010 general election. The two other states where we conducted interviews--Delaware and New Hampshire--did not provide voters with either of these alternatives, although they allowed voters to vote by absentee ballot if they provided a reason.[Footnote 32] Although seven of the nine states and the District where we conducted interviews provided voters with some option for in-person early voting, as did the 14 jurisdictions we contacted within those states, not all characterized their process as early voting. Five states-- California, Illinois, Louisiana, Maryland, and Texas--as well as the District called their process "early voting," but North Carolina called it "one-stop absentee voting" and Wisconsin called it "in- person absentee voting." Moreover, implementation and characteristics of early voting also varied among the seven states and, in some cases, among the jurisdictions within a state. * Method of voting. In three of the seven states (California, North Carolina, and Wisconsin) where we conducted interviews, voters were allowed to cast their vote in person by using vote-by-mail or absentee ballots during a specified period prior to Election Day. In these states, voters applied for an absentee or vote-by-mail ballot when they went to vote early, received a ballot on the spot, and could then cast their ballot. In contrast, in the other four states and the District, voters cast their ballots using the method voters generally use on Election Day (i.e., DRE or optical scan). * Days of early voting. Although the length of the early voting periods ranged from 7 to 30 days in the states we contacted, five of the seven states and the District required local jurisdictions to include at least one Saturday in their early voting period, and two states allowed for some jurisdiction discretion to include weekend days. Of the 14 jurisdictions we contacted that offered an early voting period, 12 included an option for voters to vote on at least one Saturday, and 6 of those jurisdictions also included at least one Sunday. For example, jurisdictions in Maryland offered a 7-day early voting period that ended 4 days before Election Day and included Saturday, but not Sunday.[Footnote 33] On the other hand, California and Wisconsin allowed voters to cast ballots in person starting about 1 month before Election Day through Election Day, and it was up to local discretion whether to include weekends. * Hours of early voting. Although seven of the nine states where we conducted interviews included at least 1 day of the weekend in their early voting period, in some jurisdictions the hours available to vote were the same for weekdays and weekends, whereas in some cases weekend hours were fewer. Sometimes the hours varied by the week of the month. For example, Louisiana, Maryland, and the District required all of their early voting sites to be open the same hours each day--9.5, 10, and 10.5 hours, respectively--Monday through Saturday. Four states-- California, Illinois, North Carolina, and Wisconsin--allowed local jurisdiction discretion to determine the hours of operation for some or all of their early voting sites. Texas used a formula based on county population to determine the number of hours, in addition to the specific days, during which early voting sites must be open.[Footnote 34] In the two Texas jurisdictions where we conducted interviews, early voting sites were open Monday through Friday for 9 or 10 hours (depending on the county) during the first week of early voting; 12 hours the second week; 12 hours on Saturday; and 5 hours or 6 hours on Sunday (depending on the county). * Number of early voting sites. The number of sites where voters could cast their ballots early, in person, also varied among the states and local jurisdictions where we conducted interviews. For example, in North Carolina there were 297 early voting sites across 100 counties, whereas in Illinois there were 180 early voting sites across 110 counties. Half of the 14 local jurisdictions we contacted that offered early voting provided voters with a single early voting site, with the size of these jurisdictions varying in terms of both registered voter population and square miles. In the 7 jurisdictions that offered more than one early voting site, voters from any precinct in the jurisdiction could cast their ballot at any of that jurisdiction's early voting sites. * Types of early voting sites. The 14 local jurisdictions we contacted also used a variety of facilities as early voting sites. In 7 of these jurisdictions, early voting locations included county clerk or election offices, schools, libraries, and community centers, as well as mobile locations. For example, in an effort to make early voting convenient, one county in Illinois provided 30 of the 180 total early voting sites used in the state, consisting of 2 permanent sites and 28 temporary sites. The 2 permanent early voting sites were county clerk offices and the remaining 28 temporary sites included community centers, libraries, senior living communities, and grocery stores, some of which were serviced by "vote mobiles"--mobile units on wheels that moved from one location to another every few days. In contrast, in the 5 local jurisdictions we contacted in California and Wisconsin, their sole early voting site was located at the local election office. See appendix V for additional details on how the local jurisdictions we contacted implemented in-person early voting for the November 2010 general election. Most Election Officials We Interviewed Expect Greater Difficulty and Costs Associated with a Weekend Election: State and local election officials we interviewed about implementing a weekend election most often identified challenges they would anticipate facing in planning and conducting Election Day activities-- specifically, finding poll workers and polling places and securing ballots and voting equipment. Election officials told us that they expected few changes to how they register voters, conduct early voting, and provide voting with provisional ballots, but they did identify other challenges with implementing federal elections on a weekend. Most Election Officials Anticipate Finding Poll Workers for a Weekend Election Would Be Difficult and Costly: Election officials we interviewed in all nine states, the District, and all 17 local jurisdictions said they would expect more poll workers would be needed for a 2-day weekend election than for a Tuesday election and related costs would increase. Further, officials in 13 of those jurisdictions and the District expected it would be more difficult to recruit a sufficient number of poll workers for a weekend election.[Footnote 35] We reported in 2006 that even though the number of poll workers needed varies by jurisdiction, having enough qualified poll workers on Election Day is crucial to ensuring that voters are able to successfully cast a vote. Nationwide, the majority of jurisdictions rely on poll workers from past elections to meet their needs, but for each election, officials also recruit new poll workers from other sources such as high schools and colleges, local businesses and organizations, and government agencies.[Footnote 36] Election officials in three jurisdictions described how changing the day for federal elections to a weekend would negatively affect their ability to draw from the poll workers and sources they have relied on in the past.[Footnote 37] For example, election officials in one local jurisdiction said that about one-fourth of their approximately 23,000 poll workers for the 2010 general election were county employees and students. A weekend election would essentially end the incentives-- paying county employees their salary and excusing students from classes--that the jurisdiction successfully used in the past to attract them to work at the polls on a Tuesday when they would normally be at work or at school. Similarly, election officials from two other jurisdictions that are required by law to provide language assistance to certain groups of voters said that they rely on younger volunteers, such as high school students, to make up the majority of their bilingual poll workers. These officials were concerned that these poll workers would be less likely to volunteer during a weekend election because the incentives used to attract them in the past-- exemption from classes--would no longer be viable. Officials from 13 jurisdictions and the District discussed other reasons why it would be difficult to recruit poll workers on one or both days of a weekend rather than on a Tuesday. For example, officials in 6 jurisdictions--including 3 jurisdictions with experience conducting weekend elections--and the District explained that some poll workers would be less willing to work on the weekend because they attend religious services or participate in church- related activities. Similarly, officials in 7 jurisdictions said they would expect other priorities, such as family obligations, sporting events, and weddings, to keep some poll workers from volunteering to work during a weekend election. Officials from one jurisdiction said that, based on their past experience with conducting an election on a Saturday, poll worker volunteers are less likely to report to work on the morning of a weekend election than they do for a Tuesday Election Day. Further, officials from 12 jurisdictions and the District said they would expect poll workers to be less willing or able to work 2 consecutive days of a weekend election due to fatigue, noting that many poll workers are elderly. Officials from one of these jurisdictions stated that many of the 2,350 poll workers who volunteered during the 2010 general election were elderly and unlikely to have the stamina to work 2 consecutive days that could each be 14 or 15 hours long. These officials further voiced concern that poll worker fatigue can lead to increased mistakes. In contrast, election officials we interviewed in 4 local jurisdictions did not anticipate difficulties finding the poll workers that would be needed for a weekend election. According to election officials in 3 of these jurisdictions, it might be easier to recruit poll workers for a weekend than for a Tuesday because a larger pool of volunteers who work Monday through Friday might be available. In a fourth jurisdiction with experience conducting state and local elections on Saturdays, officials said that while they may need to replace some poll workers that are only able or willing to work one day of a weekend election, they would expect that the compensation they offer would be sufficient to attract the number of poll workers needed to work over a weekend. However, election officials from all 17 jurisdictions and the District stated that the costs associated with poll worker pay would increase for a 2-day election, and in all but one jurisdiction, officials anticipated such costs would at least double what they spent in the 2010 general election. In that one jurisdiction, the election official anticipated poll worker costs might increase by about half--but not double--because she expected voter activity would be spread over the course of Saturday and Sunday and, thus, she would need fewer poll workers each day than for a single-day election. Moreover, election officials from 10 of these jurisdictions noted that poll worker costs represented their greatest cost in administering the 2010 general election. For example, officials from one local jurisdiction expected the number of needed poll workers and the related costs to double for a weekend election. They added that poll worker costs were already their greatest election expense, and that such an increase would significantly affect their overall election budget. Furthermore, election officials in this state said that a weekend election would at least double the $2.6 million the state incurred to help jurisdictions pay for nearly 54,000 poll workers statewide in the 2010 general election. Given its financial constraints, these officials questioned whether the state would be able to provide these payments to jurisdictions for the second day of a weekend election. In addition, election officials in three states and 4 jurisdictions noted that they might have to increase the compensation they provide poll workers or consider paying overtime to attract a sufficient number to work during a weekend election. For example, officials from a jurisdiction with less than 20 poll workers in the 2010 general election said that their costs for poll worker pay might double or triple for a weekend election because they would expect needing more poll workers as well as needing to increase compensation to successfully recruit them. Most Election Officials Expect Difficulty and Some Increased Costs Finding Polling Places for a Weekend Election: Election officials we interviewed in 14 of the 17 local jurisdictions-- including 5 jurisdictions with experience conducting elections on a Saturday--and the District expected that at least some of the polling places they used in past elections would not be available for a weekend election, and officials in all of those jurisdictions and the District anticipated difficulty finding replacements. Local election officials are responsible for selecting and securing a sufficient number of polling places that meet basic requirements and standards that include ensuring polling places are easily accessible to all voters, including voters with disabilities. They should also have a basic infrastructure capable of supporting voting machines and be comfortable for voters and poll workers, including offering sufficient indoor space and parking.[Footnote 38] The types of facilities used as polling places varied in the jurisdictions where we conducted interviews and included public and private facilities such as places of worship, schools, government buildings, fire departments, community centers, libraries, and residential facilities. Election officials noted potential challenges associated with relying on commonly used polling places on the weekend. Of the 12 jurisdictions and the District that relied on churches or synagogues for at least some of their polling places, election officials in all but one said they would need to find other locations for a weekend election because the places of worship they have relied on as polling places for Tuesday elections are used for religious services or activities on the weekend and, thus, would not be available. For example, in 2 jurisdictions where about half of the 3,067 and 200 polling places, respectively, were churches and synagogues, election officials said that they would not expect those facilities to be available on a weekend, and it would be difficult to find replacements. In contrast, in one jurisdiction with experience conducting state and local elections on a Saturday where about 15 percent of its 127 polling places were churches, election officials said they would expect the majority of those churches to remain available as polling places for a weekend election by using areas of the church not needed for religious services. However, they anticipated that churches would need to make special parking arrangements, as church goers and voters would be competing for parking spaces. Officials from 9 jurisdictions and the District explained that other polling places, such as schools and community centers, would also be more difficult to use on the weekend because of scheduled events, such as athletic events, dances, or fairs. For example, officials from one jurisdiction with past experience conducting federal elections on a Saturday stated that they had a harder time finding enough polling places for Saturday voting because fewer locations, such as community centers, were available. Officials stated that due to conflicts that prevented the use of some facilities, some polling place locations had to change from the presidential primary to the general election in the same election year. They added that, as a result, voters had to be assigned to a different polling place for the general election which caused a problem on Election Day when some of those voters went to the wrong location. In another jurisdiction where almost 70 percent of the 249 polling places in the 2010 general election were schools, officials said they would anticipate problems using schools as weekend polling places because of activities, such as athletic events, that might compete with a weekend election for space and parking. Furthermore, they found it difficult to think of any facilities that they might be able to use as replacements. In contrast, election officials from 5 jurisdictions with past experience conducting state or local elections on Saturdays noted that they might find it easier to use schools as polling places on a weekend than a Tuesday because students would not be attending classes and having students present on Election Day when campuses are open to the public has raised security concerns for some schools and jurisdictions. Officials from 2 of these jurisdictions acknowledged that schools would still have competing activities on the weekend, but anticipated they could use a different part of the school and employ additional staff to assist with parking and traffic. Regardless of the type of facility that might be unavailable as a weekend polling place, officials in 14 jurisdictions and the District said that finding alternatives would be challenging if not impossible. In all but one of these jurisdictions, officials pointed out the difficulty in locating alternative polling places that would be accessible to voters with disabilities. For example, according to one local election official, in some precincts the only building that is accessible to voters with disabilities is a church that is already used as a polling place for Tuesday elections, but would not be available on a weekend. Officials in 4 jurisdictions and the District said that in order to provide for a sufficient number of polling places they might need to consolidate precincts, in which case some voters would likely need to travel further to vote. However, in the three smallest jurisdictions in which we held interviews, election officials said they would expect the same polling places they used in past elections to still be available if the day of federal elections were moved to a weekend. In two cases, the jurisdictions had a single polling place--a municipal building--and officials would expect to use that building for a weekend election. Officials from the third jurisdiction that had experience conducting state and local elections on Saturdays, similarly stated that a weekend election would not present a challenge with respect to polling places, and they would expect to use the same 10 facilities--mostly public buildings--as polling places regardless of the day of the week the election is held. Election officials from 13 jurisdictions--including 5 jurisdictions with experience conducting elections on a Saturday--said they would expect costs associated with polling places to increase with a weekend election. Officials in 8 jurisdictions that pay for at least some of the facilities they use as polling places anticipated rental fees would double because of the 2-day aspect of a weekend election. Other officials said they would expect at least some of the facilities that are available at no cost for a Tuesday election to charge a rental fee on the weekend to compensate for potential revenue losses by, for example, not being able to rent their spaces for weddings or other private events. For example, officials from one jurisdiction said that to replace many of their 249 polling places that would be unavailable for a weekend election, they might need to offer higher compensation to attract private facilities that have not previously served as polling places. Furthermore, officials in 11 jurisdictions stated that other costs might increase with a weekend election if facilities that are normally closed on a weekend were opened for a weekend election. This might include charges for electricity or custodial and maintenance staff, who would need to be available or on the premises. In 6 of these jurisdictions, officials stated that paying for custodial or maintenance personnel might further entail overtime pay because they would be working on a weekend. Most Election Officials Said Ensuring Overnight Security of Ballots and Equipment Would Also Be Challenging and Costly: According to election officials we interviewed in all nine states, the District, and 15 of the 17 local jurisdictions, ensuring the security of ballots and voting equipment over the Saturday night of a weekend election would be both challenging and expensive. We have previously reported that secure voting systems are essential to maintaining public confidence in the election process.[Footnote 39] EAC election management guidelines further articulate that physical security safeguards are required for all voting equipment and ballots while stored, transported, and in place at polling places on Election Day, and until the time the vote is certified. Officials we interviewed in 5 of the 7 states and the District that conducted early voting and provided security over multiple days explained that the level of planning and challenges needed for overnight security for a weekend election would be on a scale that would far surpass that of early voting due to the greater number and variety of polling places used on Election Day. For example, election officials in one state observed that for the 2010 general election, the entire state had fewer than 300 early voting sites compared to more than 2,750 polling places on Election Day, and the early voting sites were selected with the need for overnight security in mind. In contrast, Election Day polling places are precinct-based and generally selected based on factors that include availability and proximity to voters rather than overnight security. In 15 of the local jurisdictions and the District, election officials said they anticipated challenges regarding the overnight security aspect of a weekend election and described the following approaches they would envision taking to ensure the security of ballots and voting equipment: * Transporting and securing ballots at another location. Election officials in 8 jurisdictions said that to ensure the security and the integrity of the election results, they would likely have ballots transported from polling places to a secure location on the Saturday night of a weekend election and back again on Sunday morning. An election official from one jurisdiction stated that municipal law requires that deputy sheriffs pick up ballots at the polling places and bring them to the clerk's office to secure them overnight during the jurisdiction's early voting period. This official stated that the jurisdiction's elections office currently employs approximately 120 deputy sheriffs to do this on Tuesday night of Election Day, and they would likely be required to do the same on Saturday night in addition to Sunday night of a weekend election. * Safeguarding voting equipment at polling places. Officials from 10 jurisdictions and the District said that to ensure overnight security during a weekend election, they would likely hire security personnel for each polling place to safeguard voting equipment from the close of polls on Saturday night until they reopen on Sunday morning. For example, an election official in one jurisdiction explained that because some of the jurisdiction's 27 polling places are located up to 100 miles from the election office, there is not enough time between polls closing Saturday night and reopening Sunday morning to transport the voting equipment to and from each polling place and the secure county office. Thus, this official said hiring security personnel and posting them at each polling place overnight would be the only viable option to ensure the security of the equipment. Officials in 3 other jurisdictions explained that two security personnel would likely be needed at each polling place not only to secure the equipment, but to provide a check and balance and safeguard the integrity of the election results. Although these officials believed that on-site security personnel would be needed, some questioned whether a sufficient number would be available. For example, officials in one jurisdiction said that even if they were to hire every off-duty police officer in their jurisdiction, they did not think they would have enough officers to secure all of their 249 polling places over the Saturday night of a weekend election. Officials from another jurisdiction anticipated that, rather than hiring security personnel, they would likely secure the voting machines on-site in a locked room to prevent tampering, vandalism, or theft, but they would need to change the locks at all of their 23 polling places. We have previously reported that larger, diverse jurisdictions can face more challenges than smaller jurisdictions, as the complexity of administering an election and the potential for challenges increase with the number of people and places involved and the scope of activities and processes that must be conducted.[Footnote 40] This might be the case with respect to ensuring overnight security during a weekend election. For example, at one extreme, election officials in the largest jurisdiction where we held interviews said they would likely employ some combination of on-site security and transporting of ballots to ensure overnight security if elections were held over 2 days. Officials explained that in their jurisdiction, which had more than 3,000 polling places on Election Day for the 2010 general election, ensuring the chain of custody of ballots on election night involved a complex logistical operation that included transporting ballots by helicopters to an estimated 70 to 80 secure locations. Given the size of their jurisdiction and the enormity of the task, these officials said they would need to assemble a task force and devote considerable resources to determine how to address Saturday night security during a weekend election since it would involve a completely new model for them and a fundamental change in procedures. In contrast, election officials in the two smallest jurisdictions where we held interviews did not anticipate overnight security would be a challenge during a weekend election, as they use a single polling place--a municipal building--on Election Day. These officials said they would expect that ballot boxes would be secured in a safe located in the county office over the Saturday night of a weekend election, just as they are at the end of a Tuesday Election Day. They added that they might consider implementing additional security measures for a weekend election, such as having police patrol the building during the weekend, but they did not anticipate this would present a challenge or represent additional costs. In addition to presenting planning and logistical challenges, election officials in all nine states, the District, and 15 of the 17 local jurisdictions where we conducted interviews said they expected the costs associated with implementing these overnight security measures to increase the cost of a weekend election. For example, in the jurisdiction that would employ deputy sheriffs to transport the ballots to the clerk's office both nights of a weekend election, the election official said this would double the more than $210,000 in security-related costs incurred for the 2010 general election. In one of the jurisdictions where officials anticipated posting two overnight security guards at each polling place, officials estimated this would add about $100,000 to their cost of administering an election. [Footnote 41] Election Officials Expected Other Challenges with Implementing Federal Elections on a Weekend: In all 17 local jurisdictions and the District, election officials reported that they would expect few changes to how they register voters, conduct early voting, and provide voting with provisional ballots. However, election officials with whom we spoke identified other challenges related to operating voting systems and reconciling ballots in preparation for counting and certifying the total number of ballots cast over a 2-day election, as well as concerns with the effect of a weekend election on workload and the election calendar. Voting technology challenges and related costs. Election officials we interviewed in 7 of the 17 local jurisdictions discussed technology- related challenges they foresaw with using their voting systems for a 2-day weekend election, and officials from 4 of these jurisdictions said they would expect addressing this to result in significantly higher costs than for a Tuesday election. According to officials, their voting systems are designed for all voting to take place in a single day and for equipment to be closed when polling places close that night. Officials explained that, to preserve the integrity of the vote in a weekend election, they would have to leave voting machines open Saturday night where polls are closed; however, the equipment could not simply be suspended Saturday night and started up again Sunday morning for a second day of voting.[Footnote 42] Rather, once closed, the equipment would, in effect, consider the election to be over and could not record additional votes. According to officials, to conduct a second day of voting, their equipment would either need to be (1) reprogrammed by the vendor in advance of the election and recertified or (2) reprogrammed Saturday night and retested before Sunday morning, which involves a lengthy process that cannot be completed in a single night. Alternatively, they could purchase additional memory cards or even a second set of voting machines. Elections officials in the City and County of San Francisco anticipated facing such a challenge in planning for a November 2011 municipal election that was to take place on 2 days--a Saturday and the following Tuesday.[Footnote 43] In consultation with the California Secretary of State's office, they determined that their voting equipment could not be closed on Saturday night and restarted on Tuesday morning. Therefore, to address this issue, they intended to borrow voting machines from other jurisdictions and use different machines each day. However, they explained that borrowing voting equipment would not be an option if the day of general elections were moved to a weekend since every jurisdiction in the country would be using its own voting equipment on the same days. Thus, they stated that if federal elections were moved to a weekend, they would likely have to purchase a second set of voting equipment to use on Sunday at over 550 polling places, at an estimated cost of over $5.9 million. This alone would represent about 88 percent of the total costs the county incurred in administering the November 2010 general election. Officials from another jurisdiction said they anticipate their voting machines would need significant changes, including changes to software, to suspend the election Saturday night and resume it on Sunday morning--changes that the officials expected would require EAC recertification.[Footnote 44] They estimated that the recertification process could take as long as 1 year and cost the manufacturer of their voting system hundreds of thousands of dollars, some of which might be passed on to them in the form of required software upgrades. Election officials in another state that used different voting equipment said they thought their equipment could suspend voting Saturday night and resume on Sunday morning if careful steps were taken by trained poll workers or technical staff on how to temporarily turn off voting machines without closing them and ending the vote. However, they would need technical staff or poll workers with more technical skills than those they have used in the past to accomplish this without ending the entire voting process by mistake. In addition, election officials in all nine states expected other related costs, such as for technology support--either in-house or contracted--would be greater for a weekend election. They stated that cost increases would primarily be due to securing these services for a second day and potentially having to pay overtime or premium pay on a weekend. For example, based on their experience conducting nonfederal elections on a Saturday, officials from Louisiana said that they would expect to incur significant additional costs because they would need to hire more part-time election staff to load and reprogram a second set of memory cards into their electronic voting machines on Sunday morning at approximately 3,000 polling places statewide. Moreover, the state normally pays to have technology vendors on call to troubleshoot equipment-related problems at polling places on Election Day, and would anticipate these costs would at least double with a 2-day election as premium pay might be involved for a weekend. Ballot reconciliation on Saturday and Sunday nights. Election officials from six states, the District, and 12 of the 17 local jurisdictions said that they would likely need to reconcile ballots-- the process of accounting for the numbers of ballots issued, unused, and spoiled and ensuring that the number of ballots cast matches the number of voters who cast ballots--on both Saturday and Sunday night of a weekend election. Officials in three of these states and 2 of these jurisdictions anticipated challenges with having to do this on 2 consecutive nights. For example, officials from one state said that in jurisdictions that use paper ballots, reconciling them on Saturday night might be difficult because it takes more time to reconcile paper ballots than other voting methods and there might not be sufficient time to complete the process before opening the polls again on Sunday morning. Election officials from another state and 2 local jurisdictions added that the work associated with reconciling ballots both nights would lengthen what is already a long day for poll workers, contribute to their fatigue, and might result in more errors in the reconciliation process. Increased election and temporary staff workload and costs. Officials from all 17 jurisdictions and the District said that the workload of local election staff would increase with a 2-day weekend election and, in all but one of the jurisdictions, said this would significantly increase personnel costs.[Footnote 45] For example, officials from one jurisdiction that employs eight full-time and one part-time election staff said that a 2-day election would require that the staff work an additional 24 hours or more with a weekend election than a Tuesday election. Further, because staff are paid a premium for weekend overtime, the $10,500 incurred in overtime costs in the November 2010 general election would at least double. Election officials in 12 of the 13 jurisdictions and the District that used temporary workers for the 2010 general election anticipated they would either need to hire more temporary workers for a weekend election or have their temporary staff work more hours, which would also result in increased costs. Effect on election calendar. Election officials in three states, the District, and all 17 jurisdictions also noted that moving the day of federal elections to a weekend could affect certain aspects of their entire election calendar--that is, dates associated with administering elections (e.g., candidates' declarations, printing ballots, voter registration, absentee ballot deadlines, and certification of the vote). Officials in 12 jurisdictions did not anticipate this would create a particular problem in administering elections in their jurisdiction. However, a state election official in New Hampshire was concerned that a weekend election might, in effect, compel his state to move its congressional primaries earlier in the year. New Hampshire's congressional primaries take place in September-- relatively late in the primary season. According to the state official, if a weekend election resulted in congressional elections being scheduled earlier than the Tuesday Election Day, the amount of time between the state's congressional primary and Election Day would not be sufficient for election officials to create the Election Day ballot. Also, officials in 3 jurisdictions and the District noted the effect that existing absentee ballot deadlines might have on voters if the day of federal elections were changed to a weekend. These officials explained that limited weekend post office hours and concerns that the U.S. Postal Service might further reduce weekend days or hours, could result in some voters--more than with a weekday election--not mailing their absentee ballots in time to be counted. For example, election officials in the District said they would expect mailed absentee ballots would need to be postmarked no later than the Saturday of a weekend election since post offices are closed on Sunday. They anticipated that under this scenario, some ballots mailed on the weekend might not be postmarked until after the election, resulting in rejected ballots. Weekend Elections Have Not Been Studied, but Studies of Other Voting Alternatives Suggest That Voter Turnout May Not Be Strongly Affected: Limited U.S. Experience with Weekend Elections Makes Evaluating Effect on Voter Turnout Challenging: Because nationwide federal elections have never been held on a weekend and we could identify few U.S. jurisdictions that have held weekend elections for state or local offices, it is difficult to draw valid conclusions about how moving federal elections to a weekend would affect voter turnout. In principle, a persuasive analysis of weekend elections would involve comparing voter turnout in jurisdictions that had moved their elections to a weekend to turnout in similar jurisdictions that continued to hold the same type of election on a Tuesday. However, since federal law requires federal elections in the United States be held on a specific Tuesday, it is not possible to use national data to estimate whether voter turnout would be different if voting took place on a weekday or weekend without making assumptions that cannot be verified.[Footnote 46] The experiences of certain state and local jurisdictions with weekend elections, as well as the experiences of other countries, might lead to speculation about how voter turnout in a weekend election in the United States would compare to turnout elsewhere. In fact, the experiences of state, local, and foreign jurisdictions do not provide good proxies for the likely U.S. experience with weekend elections for the following reasons: * State and local elections. According to the EAC, the states of Delaware, Louisiana, and Texas have had experience holding nonfederal elections or federal primaries on Saturday. However, these states' experiences do not allow for an expedient and persuasive evaluation. Historical data on state and local elections in Delaware and Texas were not easily accessible in a reliable, electronic format for the periods before, during, and after weekend elections occurred. In addition, comparing the experiences of these three states with other states would risk confusing differences in election schedules with other unobserved differences, such as state culture or campaign mobilization efforts. Further, the many unique features of each election jurisdiction limit the usefulness of this type of analysis for predicting the national effect of weekend elections. * Elections in other countries. Although other countries have had experience conducting national elections on weekends, comparisons between the United States and these countries have limited value because of differences in election laws, requirements, and civic responsibilities.[Footnote 47] For example, Australia and Brazil, which have held federal elections during the weekend in the past 5 years, generally require all eligible citizens to participate in the election process,[Footnote 48] whereas the United States makes voting optional. Differences in turnout between U.S. elections and elections in these countries may reflect different civic responsibilities in addition to different election schedules; however, it is difficult to assess which factor is actually responsible. Several other methodological challenges exist in evaluating the effect of alternative voting methods (e.g., in-person early voting, no-excuse absentee voting, and vote by mail), including weekend voting, on voter turnout. * Voting alternatives cannot easily be evaluated using randomized controlled trials that often provide the most persuasive evidence of program effect. Jurisdictions likely would not randomly assign citizens to one set of election laws without first examining potential equal-protection-type issues. * Political representatives and voters choose to adopt voting alternatives for various reasons, which might include increasing low turnout or maintaining high turnout. Consequently, the difference in turnout between jurisdictions that have or have not adopted a particular alternative could be caused by the alternative itself or by the reasons that led the jurisdiction to adopt it. * The limited number of jurisdictions that have used a particular voting alternative, or the length of time it had been in use, limit evaluations to the elections in which these alternatives have been tried. For example, researchers have evaluated vote by mail in Oregon, Washington, and selected precincts in California, because these jurisdictions have regularly used vote by mail in recent years. * Distinguishing the effect of a voting alternative from other factors that affect turnout can be challenging. These other factors include demographic, social, and psychological differences across voters; other election practices, such as registration closing dates and distance to polling places; the intensity or closeness of a campaign; and the activities of political campaigns and the news media. For example, voters in jurisdictions with highly educated, older citizens might have higher turnout and a higher propensity to use voting alternatives designed to increase turnout. Turnout might be higher in these jurisdictions, but it is unclear whether the difference is caused by the voting alternative or by the citizen characteristics that are associated with a greater motivation to vote. Further, it is difficult to assess the effect of a specific change in election practices when more than one change is made at the same time. Thus, should states make several new changes concurrently, such as implementing voter identification requirements and allowing citizens to vote in early voting periods, it would be difficult to assess the unique effect of any one change on voter turnout. Research Finds Effect of Alternative Voting Methods on Turnout Are Small and Citizen Demographics Are More Consequential: Our review of 24 studies found that alternative voting methods have small and inconsistent effects on voter turnout, as compared to demographic differences among citizens.[Footnote 49] With the exception of vote by mail, each of the alternative voting methods we reviewed was estimated to increase or decrease turnout by no more than 4 percentage points. The studies disagreed about whether the methods would increase or decrease turnout, however, as the estimates for all methods except vote by mail varied from an increase of 2.8 percentage points to a decrease of 4 percentage points, depending on the voting method and the study, as shown in table 1.[Footnote 50] Table 1: Estimated Effects of Alternative Voting Methods on Voter Turnout: Alternative voting method[A]: In-person early voting[C]; Number of studies reviewed[B]: 8; Range of years studied: 1972 to 2008; Elections studied: Presidential and congressional; Estimated effects of using the voting method on turnout (percentage points): -2.4 to 2.8. Alternative voting method[A]: Vote by mail; Number of studies reviewed[B]: 11; Range of years studied: 1960 to 2008; Elections studied: Presidential, congressional, state, primary, and special; Estimated effects of using the voting method on turnout (percentage points): -2.7 to 10.2. Alternative voting method[A]: No-excuse absentee; Number of studies reviewed[B]: 6; Range of years studied: 1972 to 2008; Elections studied: Presidential and congressional; Estimated effects of using the voting method on turnout (percentage points): -4.0 to 2.7. Alternative voting method[A]: Vote centers; Number of studies reviewed[B]: 2; Range of years studied: 2003 to 2006; Elections studied: Presidential, congressional, and state; Estimated effects of using the voting method on turnout (percentage points): -2.0 to -0.6. Alternative voting method[A]: Longer poll hours; Number of studies reviewed[B]: 1; Range of years studied: 2000; Elections studied: Presidential; Estimated effects of using the voting method on turnout (percentage points): 1.0 to 1.7. Source: GAO analysis of studies. [A] Whereas in-person early voting, vote by mail, and no-excuse absentee voting provide voters with alternatives for voting prior to Election Day, vote centers and longer poll hours seek to make voting more convenient on Election Day. We viewed the latter as alternatives to traditional precinct-place voting and, thus, included them in our review of academic research. [B] Six of the 24 studies we reviewed assessed the effect of more than one alternative voting method. Eighteen of the 24 studies provided enough detail to identify an estimated effect on turnout in percentage points. [C] No study of early voting assessed whether turnout was higher in jurisdictions that provided weekend polling hours, even though certain jurisdictions gave voters this option in the 2010 general election for federal offices. Some portion of the overall effect of early voting might be due to the provision of weekend polling hours, but the available research did not allow us to directly quantify this effect. [End of table] The maximum estimated increase suggests that alternative voting methods other than vote by mail do not increase turnout by large amounts, contrary to the goals of these policy reforms.[Footnote 51] In contrast, the estimated effects of vote by mail were larger and less consistent, ranging from a 2.7 percentage point decrease to a 10.2 percentage point increase. The maximum effect of vote by mail decreased to 6.8 percentage points when we excluded one study whose results were challenged by another study.[Footnote 52] We were unable to identify any study that directly estimated the effect of weekend elections on voter turnout in United States elections. The 24 studies showed that citizen demographics--age, education, race, income, and residential mobility--had stronger and more consistent associations with turnout than jurisdictions' use of alternative voting methods. More specifically, the studies showed the following: * A 10 percentage point increase in the percentage of a jurisdiction's population between the ages of 35 and 54 (in one study) and 45 to 64 (in another study) increased turnout by 1 to 10 percentage points. * A 10 percentage point increase in a jurisdiction's population with 4- year college degrees increased turnout by 1 to 6 percentage points. * A 10 percentage point increase in a jurisdiction's nonwhite population decreased turnout by 2 to 11 percentage points. * A $40,000 increase in a jurisdiction's median income increased turnout by 0 to 4 percentage points. * A 10 percentage point increase in a jurisdiction's renter population--a measure of residential mobility--decreased turnout by 8 percentage points. The broader academic research on voter turnout has drawn conclusions that are consistent with those of the studies we reviewed. These studies have concluded that individual differences among citizens and electoral competition are more strongly and consistently associated with the decision to vote than interventions that seek to make voting more convenient for registered voters.[Footnote 53] As a representative example, one study concluded that the association between voter age and turnout in presidential elections from 1956 through 1988 was more than five times larger than the association between voter registration closing dates prior to Election Day and turnout.[Footnote 54] Our review found that alternative voting methods have not mobilized groups of citizens who are typically less likely to vote. Five of the 24 studies examined how the effect of alternative voting methods varied across particular groups of citizens. Four of those studies showed that the methods either did not increase turnout for citizens who were typically less likely to vote, or that the methods increased turnout for citizens who were already more likely to vote. For example, one study concluded that longer poll hours did not disproportionately benefit any demographic group, including farmers and employed people working more than 40 hours per week.[Footnote 55] Another study concluded that vote by mail methods increased turnout among citizens who were well educated, older, and more interested in political campaigns.[Footnote 56] These findings suggest that alternative voting methods are more effective at retaining existing voters than mobilizing citizens who do not vote. Similarly, our review showed that citizens who were typically more likely to vote were also more likely to take advantage of early voting when it was an option. Six of the 24 studies assessed which demographic groups were more likely to vote early. These studies showed that early voters are more likely to be older, better educated, more interested in politics, and more strongly identified with a political party, as compared to voters who used other voting methods. Because these groups of citizens are typically more likely to vote, the research suggests that alternative voting methods have been more popular among citizens who need less encouragement to vote. Election officials in the nine states and the District where we conducted interviews said that they expected moving Election Day from a Tuesday to a Saturday and Sunday would have little to no effect on total voter turnout. In four of the states, officials said that a weekend election might lead to more voters voting early or absentee, but they did not think total turnout would be affected. This view was shared by officials in states that had experience in early voting, including weekend early voting, as well as states with considerable experience in holding local elections on Saturday. Their comments are generally consistent with the studies we reviewed, which assessed the effects of alternative voting methods on turnout using larger, more- representative samples of elections, jurisdictions, and time periods. Turnout Did Not Increase during the Weekend Early Voting Period in Maryland's 2010 General Election: Our analysis of voter turnout data from the early voting period during the 2010 general election in Maryland showed that voters were not very likely to vote on the weekend days provided. Maryland offered early voting for the first time in the 2010 primary and general elections. Of the voters we analyzed, 1.1 percent cast ballots on the weekend during the early voting period when they had this option during the primary election, and 1.5 percent of voters did so during the general election.[Footnote 57] The turnout rate for the general election did not increase during weekend periods of early voting, as compared to weekday periods and Election Day. About 81 percent of voters voted in person on Election Day and about 6 percent voted by absentee ballot. A total of about 11.8 percent of voters voted in person on a weekday during the state's 7-day early voting period (the second Friday through the first Thursday prior to Election Day), and about 1.5 percent voted on the Saturday of that period.[Footnote 58] Those who voted early on Saturday were generally more likely to be members of demographic groups who, according to academic research, are typically more likely to vote--that is, those who are older, less mobile, and more politically engaged. The length of registration and prior voting experience approximate a voter's residential mobility and long-term level of political engagement, respectively. However, the youngest and least experienced voters were relatively more likely to vote on Saturday, compared to voters who were slightly older and more experienced. As shown in table 2, voters who were older than 40, had been registered for at least 10 years, and voted in at least 6 of the past 10 primary and general elections were more likely to vote on Saturday in Maryland's 2010 general election than voters in other subgroups. For example, 1.4 percent of the registrants who were older than 65 and voted, voted on Saturday, compared to 1 percent of the registrants between the ages of 25 and 39 who voted. Although this change is small on an absolute scale, it is larger when expressed as a ratio of turnout rates for the two groups--a proportional difference of 45 percent. In addition to these differences, registered Democrats were 0.4 percentage points more likely than registered Republicans to have voted on the weekend--a proportional difference of 33 percent--but 6.3 percentage points less likely to have voted at all. Table 2: Turnout and Use of Saturday Poll Hours in the 2010 Maryland General Election: Voter characteristic: Age: Younger than 25; Percent of registrants voting: 26.6%; Percent voting on Saturday, among registrants who voted: 1.31%. Voter characteristic: Age: 25-39; Percent of registrants voting: 35.2%; Percent voting on Saturday, among registrants who voted: 0.96%. Voter characteristic: Age: 40-64; Percent of registrants voting: 59.1%; Percent voting on Saturday, among registrants who voted: 1.75%. Voter characteristic: Age: 65 or older; Percent of registrants voting: 63.6%; Percent voting on Saturday, among registrants who voted: 1.39%. Voter characteristic: Years registered: Less than 2; Percent of registrants voting: 36.3%; Percent voting on Saturday, among registrants who voted: 1.87%. Voter characteristic: Years registered: 2-9; Percent of registrants voting: 38.9%; Percent voting on Saturday, among registrants who voted: 1.33%. Voter characteristic: Years registered: 10-29; Percent of registrants voting: 57.8%; Percent voting on Saturday, among registrants who voted: 1.59%. Voter characteristic: Years registered: 30 or more; Percent of registrants voting: 75.1%; Percent voting on Saturday, among registrants who voted: 1.55%. Voter characteristic: Number of elections voted, 2000-2008: 0; Percent of registrants voting: 14.2%; Percent voting on Saturday, among registrants who voted: 1.73%. Voter characteristic: Number of elections voted, 2000-2008: 2; Percent of registrants voting: 35.9%; Percent voting on Saturday, among registrants who voted: 1.14%. Voter characteristic: Number of elections voted, 2000-2008: 4; Percent of registrants voting: 59.7%; Percent voting on Saturday, among registrants who voted: 1.20%. Voter characteristic: Number of elections voted, 2000-2008: 6; Percent of registrants voting: 77.9%; Percent voting on Saturday, among registrants who voted: 1.36%. Voter characteristic: Number of elections voted, 2000-2008: 8; Percent of registrants voting: 87.7%; Percent voting on Saturday, among registrants who voted: 1.75%. Voter characteristic: Number of elections voted, 2000-2008: 10; Percent of registrants voting: 95.0%; Percent voting on Saturday, among registrants who voted: 2.56%. Voter characteristic: Party registration: Democrat; Percent of registrants voting: 52.0%; Percent voting on Saturday, among registrants who voted: 1.69%. Voter characteristic: Party registration: Other; Percent of registrants voting: 35.0%; Percent voting on Saturday, among registrants who voted: 1.28%. Voter characteristic: Party registration: Republican; Percent of registrants voting: 58.3%; Percent voting on Saturday, among registrants who voted: 1.27%. Source: GAO analysis of Maryland early voting turnout data for the November 2010 general election. Note: Turnout statistics here may not match certified vote totals for Maryland jurisdictions because we analyzed a custom database of citizens registered for the 2010 general election. [End of table] Saturday turnout was slightly higher among the youngest and least- experienced subgroups of voters, as compared to voters in the subgroups immediately above them, and the most recently registered had the highest Saturday turnout of all registration groups. Because academic research has generally found that older, less mobile, and more politically engaged citizens are more likely to vote, early weekend voting appears to have been slightly more popular among Maryland citizens who need the most encouragement to vote in the first place. However, the small size of this increase suggests that Saturday poll hours did not meaningfully increase overall turnout or draw a large number of new or infrequent voters to the polls. Apart from this group, the likelihood of voting on Saturday generally increased with age, length of registration, and prior voting experience. Appendix II describes our more-detailed statistical analysis of voter turnout in Maryland. We are sending copies of this report to interested congressional committees and the EAC. In addition, the report will be available at no charge on the GAO Web site at [hyperlink, http://www.gao.gov]. If you or your staff have any questions concerning this report, please contact me at (201) 512-8777 or jenkinswo@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs can be found on the last page of this report. Key contributors to this report are listed in appendix VI. Signed by: William O. Jenkins, Jr. Director, Homeland Security and Justice Issues: [End of section] Appendix I: Selected Characteristics of States and Local Election Jurisdictions We Contacted: States are responsible for the administration of their own elections as well as federal elections, and states regulate various aspects of elections including registration procedures, absentee voting requirements, alternative voting methods, establishment of polling places, provision of Election Day workers, testing and certification of voting equipment, and counting and certification of the vote. However, local election jurisdictions--counties and subcounty governmental units, such as cities, villages, and townships--have primary responsibility for managing, planning, and conducting elections. We conducted interviews with election officials in a nonprobability sample of nine states and the District of Columbia (District), and a nonprobability sample of 17 local jurisdictions within those states, about if and how they implemented alternative voting methods and their views on how election administration and voter turnout would likely be affected in their state or jurisdiction if the day for regularly scheduled federal elections were moved to a weekend. To obtain a range of perspectives, we selected states that varied according to, among other things, geographic region, alternative voting methods provided in federal elections, experience with voting on weekends, and the level of local government responsible for administering elections (e.g., county or township) as shown in table 3. Table 3: Selected Information about the States and the District We Contacted: State: California; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: No. State: Delaware; Alternative voting method provided in 2010 elections: Early voting[A]: No; No-excuse absentee voting[B]: No; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: Yes. Primaries held on Saturday, 1978-2006; Weekend voting offered during early voting period[E]: N/A; Local election responsibility[F]: State administration at county level; Provides same-day voter registration[G]: No; Election Day is a state holiday: Yes. State: District of Columbia; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: N/A; Provides same-day voter registration[G]: Yes; Election Day is a state holiday: No. State: Illinois; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: Yes. Special congressional election held on Saturday, 2008; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: Yes. State: Louisiana; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: No; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: Yes. All nonfederal elections held on Saturday since 1959; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: Yes. State: Maryland; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: Yes. State: New Hampshire; Alternative voting method provided in 2010 elections: Early voting[A]: No; No-excuse absentee voting[B]: No; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: N/A; Local election responsibility[F]: Subcounty; Provides same-day voter registration[G]: Yes; Election Day is a state holiday: No. State: North Carolina; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: No. State: Texas; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: No; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: Yes. Some statewide elections held on Saturday; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: County; Provides same-day voter registration[G]: No; Election Day is a state holiday: No. State: Wisconsin; Alternative voting method provided in 2010 elections: Early voting[A]: Yes; No-excuse absentee voting[B]: Yes; Vote by mail only[C]: No; Experience with voting on weekends: Experience with a weekend election[D]: No; Weekend voting offered during early voting period[E]: Yes; Local election responsibility[F]: Subcounty; Provides same-day voter registration[G]: Yes; Election Day is a state holiday: No. Sources: GAO analysis of state statutes and data. Notes: N/A = Not applicable. Some of the data are from GAO synthesis of documents provided by and interviews with state election officials. [A] In general, early voting allows voters from any precinct in the jurisdiction to cast their vote in person without providing a reason before Election Day either at one specific location or at one of several locations. [B] No-excuse absentee voting is a process that allows citizens to cast an absentee ballot, usually by mail, without providing a reason. [C] Voting by mail only is a process in which all elections are conducted solely by mail. Oregon and 38 of the 39 counties in Washington administered the 2010 general election solely by mail. [D] Indicates states that have experience in conducting a federal (primary) or state election during the weekend. [E] Indicates states that offered early voting in the November 2010 general election on a Saturday, Sunday, or both. [F] In most states, election responsibility resides at the county level, although some states have delegated election responsibility to subcounty governmental units, such as cities, villages, and townships. [G] Indicates states that allowed voters to register to vote at their polling place on Election Day. [End of table] In addition, we conducted interviews with election officials in a nonprobability sample of 17 local election jurisdictions within the nine states. We selected jurisdictions to reflect variation in factors including demographics, applicable bilingual voting assistance requirements,[Footnote 59] and voting methods used, as shown in table 4. In addition, we considered other factors specific to the jurisdiction--such as for Los Angeles County, which is the largest election jurisdiction in the United States, or for San Francisco, which had developed an implementation plan for a Saturday voting pilot program for a November 2011 municipal election--in making our selections. Table 4: Selected Information about the Local Election Jurisdictions and the District We Contacted: Local jurisdiction: Los Angeles County, California; Population: 9,785,295; Percent nonwhite population[A]: 71%; Median income as a percent of national median income[B]: 107%; Bilingual voting assistance requirements[C]: Yes--Asian (Asian-Indian, Chinese, Filipino, Japanese, Korean, Vietnamese and Other Asian-Not specified), Hispanic; Voting method(s) used[D]: Optical scan. Local jurisdiction: Modoc County, California; Population: 9,162; Percent nonwhite population[A]: 20%; Median income as a percent of national median income[B]: 66%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: Optical scan. Local jurisdiction: City and County of San Francisco, California; Population: 797,271; Percent nonwhite population[A]: 55%; Median income as a percent of national median income[B]: 136%; Bilingual voting assistance requirements[C]: Yes--Asian (Chinese), Hispanic; Voting method(s) used[D]: Optical scan. Local jurisdiction: New Castle County, Delaware; Population: 527,774; Percent nonwhite population[A]: 35%; Median income as a percent of national median income[B]: 123%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE. Local jurisdiction: District of Columbia; Population: 588,433; Percent nonwhite population[A]: 68%; Median income as a percent of national median income[B]: 110%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE & Optical Scan. Local jurisdiction: Kane County, Illinois; Population: 494,371; Percent nonwhite population[A]: 37%; Median income as a percent of national median income[B]: 133%; Bilingual voting assistance requirements[C]: Yes--Hispanic; Voting method(s) used[D]: DRE. Local jurisdiction: Shelby County, Illinois; Population: 21,854; Percent nonwhite population[A]: 2%; Median income as a percent of national median income[B]: 88%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE & Optical Scan. Local jurisdiction: Orleans Parish, Louisiana; Population: 328,669; Percent nonwhite population[A]: 71%; Median income as a percent of national median income[B]: 71%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE. Local jurisdiction: Tensas Parish, Louisiana; Population: 5,798; Percent nonwhite population[A]: 59%; Median income as a percent of national median income[B]: 59%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE. Local jurisdiction: Montgomery County, Maryland; Population: 946,172; Percent nonwhite population[A]: 47%; Median income as a percent of national median income[B]: 179%; Bilingual voting assistance requirements[C]: Yes--Hispanic; Voting method(s) used[D]: DRE. Local jurisdiction: Town of Charlestown, New Hampshire; Population: 4,872; Percent nonwhite population[A]: 2%; Median income as a percent of national median income[B]: 104%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: Hand-counted paper ballot. Local jurisdiction: City of Concord, New Hampshire; Population: 42,397; Percent nonwhite population[A]: 7%; Median income as a percent of national median income[B]: 102%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: Optical Scan. Local jurisdiction: Mecklenburg County, North Carolina; Population: 861,189; Percent nonwhite population[A]: 45%; Median income as a percent of national median income[B]: 108%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE. Local jurisdiction: Sampson County, North Carolina; Population: 62,996; Percent nonwhite population[A]: 46%; Median income as a percent of national median income[B]: 68%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: DRE & Optical Scan. Local jurisdiction: Bexar County, Texas; Population: 1,584,817; Percent nonwhite population[A]: 68%; Median income as a percent of national median income[B]: 89%; Bilingual voting assistance requirements[C]: Yes--Hispanic; Voting method(s) used[D]: DRE & Optical Scan. Local jurisdiction: Webb County, Texas; Population: 231,035; Percent nonwhite population[A]: 95%; Median income as a percent of national median income[B]: 70%; Bilingual voting assistance requirements[C]: Yes--Hispanic; Voting method(s) used[D]: DRE & Optical Scan. Local jurisdiction: City of Madison, Wisconsin; Population: 229,051; Percent nonwhite population[A]: 21%; Median income as a percent of national median income[B]: 100%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: Optical Scan. Local jurisdiction: Town of Minocqua, Wisconsin; Population: 4,935; Percent nonwhite population[A]: 2%; Median income as a percent of national median income[B]: 84%; Bilingual voting assistance requirements[C]: No; Voting method(s) used[D]: Optical Scan. Sources: Statutes, states and local jurisdictions, and GAO analysis of U.S. Census Bureau data. Notes: Bilingual voting assistance coverage determinations under sections 4(f)(4) and 203 of the Voting Rights Act of 1965, as amended, are reflected at 28 C.F.R. Part 55, Appendix (Jurisdictions Covered Under Sections 4(f)(4) and 203(c) of the Voting Rights Act of 1965, as amended) with the most recent section 203 coverage determinations published in October 2011 at 76 Fed. Reg. 63,602-63,607 (2011); state or local jurisdiction election data are from their websites; U.S. Census Bureau data are from the American Community Survey. [A] Based on analysis of the U.S. Census Bureau's American Community Survey data. Indicates the percentage of the population that is either (1) Hispanic and any race, or (2) non-Hispanic and any race other than white. [B] Represents the median household income of the jurisdiction as a percentage of the national median household income. [C] Indicates whether or not a jurisdiction is required under section 203 or 4(f)(4) of the Voting Rights Act of 1965, as amended, to provide bilingual election materials and assistance and, if so, the language minority group(s) covered. [D] In general, there are five methods for casting votes: paper ballots, lever machines, punch cards, optical scan, and direct recording electronic (DRE) machines. An optical scan voting system is composed of computer-readable ballots, appropriate marking devices, privacy booths, and a computerized tabulation machine. Voters record their choices using an appropriate writing instrument to fill in boxes or ovals or to complete an arrow next to the candidate's name or the issue. Voters make their selection on DREs by pressing a button or touching the screen next to the candidate's name or ballot issue. When they are finished, they cast their votes by pressing a final "vote" button on the machine or screen. [End of table] [End of section] Appendix II: Analysis of Weekend Voting in the 2010 Maryland State Elections: The state of Maryland provided its citizens the option of in-person early voting for the first time in the 2010 primary and general elections. Polls were open for early voting on a total of 6 days, beginning the second Friday prior to Election Day (September 14 or November 2, respectively, for the primary and general elections) and extending through the first Thursday prior to Election Day. Early voting hours were provided on Saturday, but not on Sunday, of each 7- day early voting period. State statute required counties to establish early voting centers, with the number of early voting locations based on the county's number of registered voters. Each county had at least one location, plus three to five additional locations if they had more than 150,000 registered voters[Footnote 60]. Early voting hours were the same across counties, beginning at 10:00 a.m. and ending at 8:00 p.m. each day. Maryland's experience with early voting allowed us to analyze how voters used weekend poll hours when they were available. Voter registration and turnout data in Maryland are sufficiently detailed and reliable to allow for statistical analysis of citizens who were registered for the 2010 general election. This appendix presents our analysis of (1) whether the turnout rate during the early voting period was higher or lower on Saturday as compared to weekdays and (2) which groups of citizens used weekend poll hours in the 2010 general election. Specifically, we assessed whether citizens who belonged to groups that typically vote less frequently, such as younger and more- recently registered voters, were more likely to use weekend poll hours. While our analysis describes the use of weekend poll hours, it does not seek to estimate the causal effect of providing these voting methods or holding Election Day on Saturday and Sunday. Turnout Did Not Substantially Increase during Weekend Poll Hours: Our analysis of voter turnout data showed that only 1.5 percent of voters used Saturday poll hours during the early voting period of the 2010 general election.[Footnote 61] To further examine how the turnout rate changed between the weekend and weekday periods, we analyzed the voting times for early voters. According to state officials, all counties in Maryland used the same computerized voter registration and election administration system in 2010, which recorded the date and time when each voter received a ballot. By estimating the turnout rate within small intervals during the early voting period, we assessed whether turnout meaningfully changed between the weekday and weekend periods. As shown in figure 3, the proportion of Maryland voters--categorized into groups by age, length of registration, and participation in prior elections--who cast ballots on a certain "poll day" during the early voting period did not substantially increase on Saturday. In our analysis, a poll day is a 24-hour period when the polls were open during the early voting period. It equals the calendar days prior to Election Day when citizens were able to vote minus the subsequent time when the polls were closed.[Footnote 62] For example, figure 3 shows that the first citizen to receive a ballot when the polls opened on Saturday of the early voting period voted 2.9 poll days prior to Election Day, even though Saturday, October 23, was the 10th calendar day prior to Election Day on Tuesday, November 2. Figure 3: Early Voting in Maryland's November 2010 General Election: [Refer to PDF for image: 3 multiple line graphs] Graphs depict early voting in poll days prior to election as follows: Proportion Voting Prior to Election day by Age; Age 18-24l Age 25-39; Age 40-64; Age 65+. Proportion Voting Prior to Election day by Length or registration; 0-1 years registered; 2-9 years registered; 10-29 years registered; 30+ years registered. Proportion Voting Prior to Election day by Participation in prior elections; 0 prior elections voted; 1-3 prior elections voted; 4-6 prior elections voted; 7+ prior elections voted. Source: GAO analysis of Maryland early voting turnout data for the November 2010 general election. Note: A "poll day" is a 24-hour period when polls were open for early voting before Election Day. The dashed vertical lines in the figure show the start of each new calendar day, with the 1 day of Saturday voting in the early voting period labeled in the second interval. [End of figure] In figure 3, the turnout rate is roughly equal to the slope of the curves plotted, with a steeper curve meaning that more voters cast ballots at a certain time.[Footnote 63] If the appeal of voting on Saturday were much higher than voting on a weekday, we would expect a sharp, discontinuous change in the turnout rate between Friday and Saturday and Saturday and Monday, respectively. The data show no such pattern. The turnout rate was approximately the same when the polls closed on Friday afternoon as when they opened on Saturday morning. Turnout increased at the start of the next weekday period on Monday, and then it steadily increased until Election Day. This suggests that voters were not relatively more eager to take advantage of Saturday voting as compared to weekday voting. Voters from different demographic groups used the Saturday voting period somewhat differently. Figure 3 plots turnout separately by age, length of registration (a proxy for residential mobility), and the number of statewide primary and general elections since 2000 in which the citizen voted. These are three variables typically associated with greater turnout. If citizens with different propensities to vote were equally likely to take advantage of weekend hours, the turnout curves should be roughly parallel on Saturday and across the boundaries between the weekday and weekend. In fact, the least-experienced citizens--those who were between the ages of 17 and 24, registered for no more than 1 year, or had never voted before--turned out at a slightly higher rate on Saturday than the other groups, and their turnout increased somewhat discontinuously between Friday and Saturday. However, the small size of this increase suggests that Saturday poll hours did not meaningfully increase overall turnout or draw a large number of new or infrequent voters to the polls. In subsequent days, however, the trend was reversed, with more- experienced citizens voting at higher rates until the middle of the final day of early voting. Less-experienced voters lagged behind and were much more likely to vote on the last day of early voting, particularly in the last half of the day. An alternative way to analyze how voters used early poll hours is to estimate the "hazard" of voting prior to Election Day for those citizens who voted early. In this case, we interpret the hazard to be the number of voters casting ballots within a small interval of time, or the instantaneous turnout rate. A larger hazard at a certain time in the early voting period means that a citizen had a higher probability of voting in a small interval around that time. We estimated the hazard function for citizens who voted during the early voting period for each of the subgroups in figure 3, using Epinechnikov kernel-smoothed estimates of the Kaplan-Meier hazard function. Estimates of hazard functions for each age group generally support similar conclusions as the simpler distributions of voting times shown in figure 3--the turnout rate was low until the fourth day of early voting, on Tuesday. At that time, the rates began to increase for all age groups, and then increased at a much larger rate on the final two days of early voting, on Wednesday and Thursday. There is no evidence that turnout increased by a large amount on Saturday. However, turnout for citizens younger than 25 increased by a slightly larger amount between Friday and Saturday than it did for citizens between the ages of 25 and 64, consistent with the evidence in figure 3. Multivariate Statistical Analysis of Saturday Voting across Groups Produces Similar Results: In order to describe the patterns in figure 3 more precisely, we used several statistical methods to estimate how turnout and the use of Saturday voting varied across groups of citizens with different characteristics. These methods allowed us to estimate the association between a certain characteristic and outcomes of interest, such as age and prior turnout, while holding constant other characteristics, such as the length of registration. In particular, we estimated two types of statistical models of the registrant's decision to vote and the voter's decision to vote on Saturday. The models were given by: E(Votei); =; (ß0 + Agei ß1 + Tenurei ß2 + Sexi ß3 + Partyi ß4 + Countyi ß5). E(Weekendi | Votei = 1); =; (0 + Agei 1 + Tenurei 2 + Sexi 3 + Partyi 4 + Countyi 5), where Votei is a Bernoulli random variable indicating whether registrant i voted; is the logistic function; Weekendi indicates whether a voter voted on Saturday; and the remaining terms are vectors of parameters and indicator covariates as specified in table 5. (Countyi is a vector of indicators for each county.) [Footnote 64] Table 5: Turnout and Use of Weekend Poll Hours in 2010 Maryland General Election, by Demographic Group: Variable: Age; Age: Younger than 25; Raw data: Probability of registrant voting: 26.6%; Probability of voting on Saturday, given that registrant voted: 1.31%; Model estimates: Probability of registrant voting: 34.0%; Probability of voting on Saturday, given that registrant voted: 1.20%. Age: 25-39; Raw data: Probability of registrant voting: 35.2%; Probability of voting on Saturday, given that registrant voted: 0.96%; Model estimates: Probability of registrant voting: 40.6%; Probability of voting on Saturday, given that registrant voted: 1.03%. Age: 40-64; Raw data: Probability of registrant voting: 59.1%; Probability of voting on Saturday, given that registrant voted: 1.75%; Model estimates: Probability of registrant voting: 57.9%; Probability of voting on Saturday, given that registrant voted: 1.80%. Age: 65 or older; Raw data: Probability of registrant voting: 63.6%; Probability of voting on Saturday, given that registrant voted: 1.39%; Model estimates: Probability of registrant voting: 54.8%; Probability of voting on Saturday, given that registrant voted: 1.37%. Variable: Years registered: Years registered: Less than 2; Raw data: Probability of registrant voting: 36.3%; Probability of voting on Saturday, given that registrant voted: 1.87%; Model estimates: Probability of registrant voting: 45.7%; Probability of voting on Saturday, given that registrant voted: 2.23%. Years registered: 2-9; Raw data: Probability of registrant voting: 38.9%; Probability of voting on Saturday, given that registrant voted: 1.33%; Model estimates: Probability of registrant voting: 43.5%; Probability of voting on Saturday, given that registrant voted: 1.45%. Years registered: 10-29; Raw data: Probability of registrant voting: 57.8%; Probability of voting on Saturday, given that registrant voted: 1.59%; Model estimates: Probability of registrant voting: 53.6%; Probability of voting on Saturday, given that registrant voted: 1.54%. Years registered: 30 or more; Raw data: Probability of registrant voting: 75.1%; Probability of voting on Saturday, given that registrant voted: 1.55%; Model estimates: Probability of registrant voting: 68.7%; Probability of voting on Saturday, given that registrant voted: 1.58%. Variable: Number of elections voted, 2000-2008: Number of elections voted, 2000-2008: 0; Raw data: Probability of registrant voting: 14.2%; Probability of voting on Saturday, given that registrant voted: 1.73%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 1; Raw data: Probability of registrant voting: 26.6%; Probability of voting on Saturday, given that registrant voted: 1.05%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 2; Raw data: Probability of registrant voting: 35.9%; Probability of voting on Saturday, given that registrant voted: 1.14%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 3; Raw data: Probability of registrant voting: 48.4%; Probability of voting on Saturday, given that registrant voted: 1.05%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 4; Raw data: Probability of registrant voting: 59.7%; Probability of voting on Saturday, given that registrant voted: 1.20%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 5; Raw data: Probability of registrant voting: 71.6%; Probability of voting on Saturday, given that registrant voted: 1.32%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 6; Raw data: Probability of registrant voting: 77.9%; Probability of voting on Saturday, given that registrant voted: 1.36%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 7; Raw data: Probability of registrant voting: 83.9%; Probability of voting on Saturday, given that registrant voted: 1.48%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 8; Raw data: Probability of registrant voting: 87.7%; Probability of voting on Saturday, given that registrant voted: 1.75%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 9; Raw data: Probability of registrant voting: 91.8%; Probability of voting on Saturday, given that registrant voted: 1.98%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Number of elections voted, 2000-2008: 10; Raw data: Probability of registrant voting: 95.0%; Probability of voting on Saturday, given that registrant voted: 2.56%; Model estimates: Probability of registrant voting: N/A; Probability of voting on Saturday, given that registrant voted: N/A. Variable: Party Registration: Party Registration: Democrat; Raw data: Probability of registrant voting: 52.0%; Probability of voting on Saturday, given that registrant voted: 1.69%; Model estimates: Probability of registrant voting: 52.1%; Probability of voting on Saturday, given that registrant voted: 1.71%. Party Registration: Other; Raw data: Probability of registrant voting: 35.0%; Probability of voting on Saturday, given that registrant voted: 1.28%; Model estimates: Probability of registrant voting: 40.0%; Probability of voting on Saturday, given that registrant voted: 1.30%. Party Registration: Republican; Raw data: Probability of registrant voting: 58.3%; Probability of voting on Saturday, given that registrant voted: 1.27%; Model estimates: Probability of registrant voting: 54.8%; Probability of voting on Saturday, given that registrant voted: 1.31%. Source: GAO analysis of Maryland early voting turnout data for the November 2010 general election. N/A = Not applicable. [End of table] To assess marginal effects, we estimated the in-sample mean predicted probabilities for each level of each covariate (though table 5 includes estimates only for the covariates of interest). We estimated robust standard errors of the parameters and predicted probabilities but do not report them here for simplicity. The standard errors were no more than 5 percent of the estimated probabilities, which partially reflects sample sizes of 1,857,675 for the model of turnout and 927,774 for the model of weekend voting. For ease of computation, we estimated the models on a 50 percent simple random sample of the population of registrants. The model estimates support the patterns in the raw data. Relatively fewer young citizens chose to vote, and most of those who did were not more likely to have voted on Saturday. Similarly, the most recently registered voters were also less likely to vote; however, in contrast, they were more likely to vote on Saturday, holding constant differences associated with age. On an absolute scale, however, few voters used Saturday poll hours, and a far greater proportion of less- experienced voters either did not vote, voted late in the early voting period, or waited until Election Day. Specifically, although our model estimates that no more than 2.2 percent of any subgroup of voters cast their ballots on Saturday, holding constant other group memberships, older voters were relatively more likely to do so than younger voters. The adjusted probability of voting on Saturday for voters who were between the ages of 40 and 64 was 1.8 percentage points, as compared to 1.2 percentage points for voters who were younger than 25--a difference of 50 percent expressed as a ratio. The analogous probabilities for voters registered less than 2 years ago and between 2 and 9 years ago were 2.2 and 1.5 percentage points, respectively, or a difference of 47 percent. The probability of voting on Saturday was slightly lower among citizens at least 65 years old, as compared to citizens between the ages of 40 and 64. Less-experienced citizens were much less likely to have voted in the first place. Citizens younger than 25 were 37 percentage points less likely to vote than citizens 65 and older. Similarly, citizens who first registered within the past 2 years were 39 percentage points less likely to vote than citizens who had been registered for 30 years or more. The national experience with holding regular elections on Saturday and Sunday might differ in meaningful ways from Maryland's experience with allowing early voting on the weekend. Maryland citizens are not necessarily representative of the nation, and in 2010 the state's early voting program was in its first year of operation. Voters may use weekend poll hours differently as they continue to learn about this option. Moreover, early voter behavior may not resemble voter behavior in elections where Election Day falls on Saturday and Sunday. In the latter system, political campaigns and the news media may increase voter awareness of weekend poll hours, and voters would not be forced to choose between voting on the weekend and voting before the political campaigns have ended. Despite these limitations, our analysis suggests that relatively few voters used weekend poll hours when they were offered in the 2010 Maryland general election, and that most of the citizens in subgroups typically less likely to vote did not turn out at vastly higher rates during this period. If voters' behavior can accurately reveal their preferences for different voting methods, the demand for weekend poll hours appeared to be modest in this election. [End of section] Appendix III: Alternative Voting Methods Provided in 50 States and the District for the 2004 and 2010 November General Elections: The number of states providing alternative voting methods--that is, in- person early voting and no-excuse absentee voting--has increased, as shown in figure 4. Specifically, in 2006, on the basis of results from a survey of 50 states and the District of Columbia (District), we reported that 24 states and the District required or allowed in-person early voting and 21 states allowed or required no-excuse absentee voting by mail in the November 2004 general election.[Footnote 65] For the November 2010 general election, 33 states and the District provided in-person early voting and 29 states and the District provided no-excuse absentee voting by mail. Figure 4: In-person Early Voting and No-excuse Absentee Voting in 50 States and the District for the 2004 and 2010 General Elections: [Refer to PDF for image: 4 illustrated U.S. maps] 2004 In-person early voting: Alaska: Arizona: Arkansas: California: Colorado: District of Columbia: Florida: Georgia: Hawaii: Idaho: Iowa: Kansas: Louisiana: Montana: Nebraska: Nevada: New Mexico: North Carolina: North Dakota: Oklahoma: Tennessee: Texas: Utah: Vermont: West Virginia: 2010 In-person early voting: Alaska: Arizona: Arkansas: California: Colorado: District of Columbia: Florida: Georgia: Hawaii: Idaho: Illinois: Indiana: Iowa: Kansas: Louisiana: Maine: Maryland: Montana: Nebraska: Nevada: New Jersey: New Mexico: North Carolina: North Dakota: Ohio: Oklahoma: South Dakota: Tennessee: Texas: Utah: Vermont: West Virginia: Wisconsin: Wyoming: 2004 No-excuse absentee voting by mail: California: Colorado: District of Columbia: Florida: Hawaii: Idaho: Iowa: Kansas: Maine: Montana: Nebraska: Nevada: New Mexico: North Carolina: North Dakota: Oklahoma: South Dakota: Utah: Vermont: Washington: Wisconsin: Wyoming: 2010 No-excuse absentee voting by mail: Alaska: Arizona: California: Colorado: Delaware: District of Columbia: Florida: Georgia: Hawaii: Idaho: Illinois: Iowa: Kansas: Maine: Maryland: Montana: Nebraska: Nevada: New Jersey: New Mexico: North Carolina: North Dakota: Ohio: Oklahoma: Oregon: South Dakota: Utah: Vermont: Washington: Wisconsin: Wyoming: Source: GAO, and GAO analysis of state statutes (data); MapQuest (map). Notes: Data for 2004 are from GAO-06-451SP and data for 2010 are from state statutes. [End of figure] [End of section] Appendix IV: Selected Details of Early and No-Excuse Absentee Voting for the 2010 General Election in States We Contacted: Of the nine states and the District of Columbia (District) we contacted, seven states and the District provided early voting. Of those seven states, five states and the District provided both early voting and no-excuse absentee voting. Two of the nine states where we conducted interviews--Delaware and New Hampshire--did not provide voters with either of these alternatives, although they allowed voters to vote by absentee ballot if they provided a reason. Table 6 provides selected details on how early and no-excuse absentee voting were implemented during the November 2010 general election. Table 6: Selected Details of Early and No-Excuse Absentee Voting Implementation for the November 2010 General Election in States We Contacted: State: California; Early voting: Year early voting began[A]: 1978; Early voting calendar days: 29; Include Saturday or Sunday: Up to local discretion; County determines specific days and hours; Number of locations: 58 locations (one per county); Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Received by close of polls on Election Day. State: District of Columbia; Early voting: Year early voting began[A]: 2010; Early voting calendar days: 14; Include Saturday or Sunday: Yes, two Saturdays; District determines specific days and hours; Early voting: Number of locations: 5 locations across 143 precincts; Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Postmarked on or by Election Day. State: Illinois; Early voting: Year early voting began[A]: 2006; Early voting calendar days: 18; Include Saturday or Sunday: Yes, state requires permanent locations to be open weekends and holidays. Days for temporary locations are up to local discretion; Number of locations: 180 locations across 110 counties; Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Postmarked by the day prior to Election Day. State: Louisiana; Early voting: Year early voting began[A]: 2006; Early voting calendar days: 8; Include Saturday or Sunday: Yes, Saturday. State required the number of days and hours; Number of locations: 85 locations across 64 parishes; Absentee voting: No-excuse absentee voting: No; Absentee ballot deadline[B]: Received by 4:30 p.m. the day prior to Election Day. State: Maryland; Early voting: Year early voting began[A]: 2010; Early voting calendar days: 7; Include Saturday or Sunday: Yes, Saturday; State required the number of days and hours; Number of locations: 46 locations across 24 counties; Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Postmarked on or by Election Day. State: North Carolina; Early voting: Year early voting began[A]: 2000; Early voting calendar days: 17; Include Saturday or Sunday: Yes, state requires early voting be available the last Saturday prior to Election Day. Other days and hours up to local discretion; Number of locations: 297 locations across 100 counties; Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Postmarked on or by Election Day. State: Texas; Early voting: Year early voting began[A]: 1988; Early voting calendar days: 12; Include Saturday or Sunday: Yes, Saturday and Sunday; State requirement based on county population size; Number of locations: 938 locations across 254 counties; Absentee voting: No-excuse absentee voting: No; Absentee ballot deadline[B]: Received by close of polls on Election Day. State: Wisconsin; Early voting: Year early voting began[A]: 2000; Early voting calendar days: 30; Include Saturday or Sunday: Up to local discretion; Municipality determines specific days and hours; Number of locations: 1,850 locations (one per municipality); Absentee voting: No-excuse absentee voting: Yes; Absentee ballot deadline[B]: Received by close of polls on Election Day. Source: GAO analysis of state statutes and data as well as GAO synthesis of state data. Notes: GAO synthesized data from documents provided by and interviews with state election officials. [A] Represents the year early voting was first implemented. [B] Indicates when domestic absentee ballots are due. Variation exists among states on how absentee ballots are delivered, such as by mail, in person, or by facsimile, among others. Absentee ballots must be either received or postmarked on or by Election Day. [End of table] [End of section] Appendix V: Selected Details of Early Voting for the 2010 General Election in Local Jurisdictions We Contacted: Of the 17 local jurisdictions and the District of Columbia (District) we contacted, 14 jurisdictions and the District provided in-person early voting. Table 7 provides selected details regarding how early voting was implemented during the November 2010 general election. Table 7: Selected Details of Early Voting Implementation for the November 2010 General Election in Local Jurisdictions We Contacted: Jurisdiction: Los Angeles County, California; Number of locations: 1 (election/main office); Include Saturday or Sunday: Yes. Both Saturdays and Sundays were included during the last 2 weekends prior to Election Day; Early voting hours: Election/main office: 9 hours on weekdays; 8 hours on weekends; Other locations[A]: N/A. Jurisdiction: Modoc County, California; Number of locations: 1 (election/main office); Include Saturday or Sunday: No; Early voting hours: Election/main office: 8 hours on weekdays; Other locations[A]: N/A. Jurisdiction: San Francisco County, California; Number of locations: 1 (election/main office); Include Saturday or Sunday: Yes. Both Saturdays and Sundays were included during the last 2 weekends prior to Election Day; Early voting hours: Election/main office: 9 hours on weekdays; 6 hours on weekends; Other locations[A]: N/A. Jurisdiction: District of Columbia; Number of locations: 5; Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 10.5 hours on weekdays and Saturday[B]; Other locations[A]: Same as main office. Jurisdiction: Kane County, Illinois; Number of locations: 30; Include Saturday or Sunday: Yes. Included Saturdays and Sundays; Early voting hours: Election/main office: For permanent locations, 8 or more hours on weekdays and 3 hours on weekends. For temporary locations, up to local discretion; Other locations[A]: Hours varied. Jurisdiction: Shelby County, Illinois; Number of locations: 1 (election/main office); Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 8 hours on weekdays; 4 hours on Saturday; Other locations[A]: N/A. Jurisdiction: Orleans Parish, Louisiana; Number of locations: 3; Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 9.5 hours on weekdays and Saturday; Other locations[A]: Same as main office. Jurisdiction: Tensas Parish, Louisiana; Number of locations: 1 (election/main office); Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 9.5 hours on weekdays and Saturday; Other locations[A]: N/A. Jurisdiction: Montgomery County, Maryland; Number of locations: 5; Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 10 hours on weekdays and Saturday; Other locations[A]: Same as main office. Jurisdiction: Mecklenburg County, North Carolina; Number of locations: 16; Include Saturday or Sunday: Yes. Included 2 Saturdays and 1 Sunday; Early voting hours: Election/main office: 9-11 hours on weekdays; 3 hours on Saturdays and Sunday; Other locations[A]: 8 hours on weekdays; 3 hours on Saturdays and Sunday. Jurisdiction: Sampson County, North Carolina; Number of locations: 4; Include Saturday or Sunday: Yes. Included 3 Saturdays; Early voting hours: Election/main office: 11 hours on weekdays; 5-7 hours on Saturdays; Other locations[A]: 6 hours on weekdays; 5-7 hours on Saturdays. Jurisdiction: Bexar County, Texas; Number of locations: 40; Include Saturday or Sunday: Yes. Included Saturday and Sunday; Early voting hours: Election/main office: 10-12 hours on weekdays; 12 hours on Saturday; 6 hours on Sunday; Other locations[A]: Same as main office. Jurisdiction: Webb County, Texas; Number of locations: 9; Include Saturday or Sunday: Yes. Included Saturday and Sunday; Early voting hours: Election/main office: 9-12 hours on weekdays; 12 hours on Saturday; 5 hours on Sunday; Other locations[A]: Hours varied with temporary and mobile locations. Jurisdiction: City of Madison, Wisconsin; Number of locations: 1 (election/main office); Include Saturday or Sunday: Yes. Included Saturday; Early voting hours: Election/main office: 8.5-11 hours on weekdays; 3 hours on Saturday; Other locations[A]: N/A. Jurisdiction: Town of Minocqua, Wisconsin; Number of locations: 1 (election/main office); Include Saturday or Sunday: No; Early voting hours: Election/main office: 8 hours on weekdays; Other locations[A]: N/A. Source: GAO interviews of local election jurisdiction officials. Note: N/A = not applicable. [A] Includes temporary or satellite locations, such as schools, libraries, and community centers, among others, that allowed early voting. [B] Early voting was provided for 8 hours and 15 minutes at the main election office the last day of the early voting period (the day before Election Day). [End of table] [End of section] Appendix VI: GAO Contact and Staff Acknowledgments: GAO Contact: William O. Jenkins, Jr., (202) 512-8777 or jenkinswo@gao.gov: Staff Acknowledgments: In addition to the contact named above, Mary Catherine Hult, Assistant Director; David Alexander; Josh Diosomito; Geoffrey Hamilton; Lara Miklozek; Hugh Paquette; Jeff Tessin; and Lori Weiss made key contributions to this report. [End of section] Bibliography: Bergman, Elizabeth, Philip Yates, and Elaine Ginnold. "How Does Vote by Mail Affect Voters? A natural experiment examining individual-level turnout." The PEW Center on the States, Make Voting Work project. Accessed May 19, 2011. [hyperlink, http://www.pewcenteronthestates.org/report_detail.aspx?id=58252]. Berinsky, Adam J., Nancy Burns, and Michael W. Traugott. "Who Votes By Mail? A Dynamic Model of the Individual-Level Consequences of Voting- By-Mail Systems." The Public Opinion Quarterly, vol. 65 (2001): 178- 197. Burden, Barry C., David T. Canon, Kenneth R. Mayer, and Donald P. Moynihan. "Election Laws, Mobilization, and Turnout: The Unanticipated Consequences of Election Reform," April 12, 2011. Social Science Research Network eLibrary. Accessed May 19, 2011. [hyperlink, http://ssrn.com/abstract=1690723]. Fitzgerald, Mary. "Greater Convenience but Not Greater Turnout--the Impact of Alternative Voting Methods on Electoral Participation in the United States." American Politics Research, vol. 33 (2005): 842-867. Giammo, Joseph D., and Brian J. Brox. "Reducing the Costs of Participation: Are States Getting a Return on Early Voting?" Political Research Quarterly, vol. 63 (2010): 295-303. Gronke, Paul, and Daniel Krantz Toffey, "The Psychological and Institutional Determinants of Early Voting." Journal of Social Issues, vol. 64 (2008): 503-524. Gronke, Paul, Eva Galanes-Rosenbaum, and Peter A. Miller. "Early Voting and Turnout." PS: Political Science and Politics, vol. 40 (2007): 639-645. Gronke, Paul, Eva Galanes-Rosenbaum, and Peter A. Miller. "Early Voting and Voter Turnout." In Democracy in the States: Experiments in Election Reform. Ed. Bruce E. Cain, Todd Donovan, and Caroline J. Tolbert. Washington, D.C.: Brookings Institution, 2008. Gronke, Paul, and Peter A.M. Miller. "Voting by Mail and Turnout: A Replication and Extension." Paper presented at the 2007 Annual Meeting of the American Political Science Association, Chicago, Ill. Hanmer, Michael J., and Michael W. Traugott. "The Impact of Voting by Mail on Voter Behavior." American Politics Research, vol. 32 (2004): 375-405. Hill, David, and Michael D. Martinez. "The Interactive Effects of Electoral Reform, Competitiveness and Age on Turnout." Paper presented at the 2008 Annual Meeting of the American Political Science Association, Boston, Mass. Juenke, Eric Gonzalez, and Juliemarie Shepherd. "Vote Centers and Voter Turnout." In Democracy in the States: Experiments in Election Reform. Ed. Bruce E. Cain, Todd Donovan, and Caroline J. Tolbert. Washington, D.C.: Brookings Institution, 2008. Karp, Jeffrey A., and Susan A. Banducci. "Going Postal: How All-Mail Elections Influence Turnout." Political Behavior, vol. 22 (2000): 223- 239. Kousser, Thad and Megan Mullin. "Does Voting by Mail Increase Participation? Using Matching to Analyze a Natural Experiment." Political Analysis, vol. 15 (2007): 428-445. Leighley, Jan E., and Jonathan Nagler. "Electoral Laws and Turnout, 1972-2008." Paper presented at the 4th Annual Conference on Empirical Legal Studies, Los Angeles, Calif., November 2009. Magleby, David B. "Participation in Mail Ballot Elections." The Western Political Quarterly, vol. 40 (1987): 79-91. Miller, Peter A., and Neilan S. Chaturvedi. "Get Out the Early Vote: Minority Use of Convenience Voting in 2008." Paper presented at the 2010 Annual Meeting of the American Political Science Association, Washington, D.C. Miller, Peter A.M., and Paul Gronke. "The Effect of Voting by Mail in Washington: 1960-2008." Portland, Ore.: The Early Voting Information Center, Reed College. Accessed May 19, 2011. [hyperlink, http://www.earlyvoting.net/research]. Patterson, Samuel C. and Gregory A. Caldeira. "Mailing in the Vote: Correlates and Consequences of Absentee Voting." American Journal of Political Science, vol. 29 (1985): 766-788. Southwell, Priscilla L., and Justin I. Burchett. "The Effect of All- Mail Elections on Voter Turnout." American Politics Quarterly, vol. 28 (2000): 72-79. Stein, Robert M. "Early Voting." The Public Opinion Quarterly, vol. 62 (1998): 57-69. Stein, Robert M., and Greg Vonnahme. "Engaging the Unengaged Voter: Voter Centers and Voter Turnout." The Journal of Politics, vol. 70 (2008): 487-497. Stein, Robert M., and Patricia A. Garcia-Monet. "Voting Early but Not Often." Social Science Quarterly, vol. 78 (1997): 657-671. Wolfinger, Raymond E., Benjamin Highton, and Megan Mullin. "How Postregistration Laws Affect the Turnout of Citizens Registered to Vote." State Politics and Policy Quarterly, vol. 5 (2005): 1-23. [End of section] Related GAO Products: Elections: DOD Can Strengthen Evaluation of Its Absentee Voting Assistance Program. [hyperlink, http://www.gao.gov/products/GAO-10-476]. Washington, D.C.: June 17, 2010. Elderly Voters: Information on Promising Practices Could Strengthen the Integrity of the Voting Process in Long-term Care Facilities. [hyperlink, http://www.gao.gov/products/GAO-10-6]. Washington, D.C.: November 30, 2009. Voters with Disabilities: Additional Monitoring of Polling Places Could Further Improve Accessibility. [hyperlink, http://www.gao.gov/products/GAO-09-941]. Washington, D.C.: September 30, 2009. Voters With Disabilities: More Polling Places Had No Potential Impediments Than in 2000, but Challenges Remain. [hyperlink, http://www.gao.gov/products/GAO-09-685]. Washington, D.C.: June 10, 2009. Elections: States, Territories, and the District Are Taking a Range of Important Steps to Manage Their Varied Voting System Environments. [hyperlink, http://www.gao.gov/products/GAO-08-874]. Washington, D.C.: September 25, 2008. Elections: Federal Program for Certifying Voting Systems Needs to Be Further Defined, Fully Implemented, and Expanded. [hyperlink, http://www.gao.gov/products/GAO-08-814]. Washington, D.C.: September 16, 2008. Bilingual Voting Assistance: Selected Jurisdictions' Strategies for Identifying Needs and Providing Assistance. [hyperlink, http://www.gao.gov/products/GAO-08-182]. Washington, D.C.: January 18, 2008. Elections: Action Plans Needed to Fully Address Challenges in Electronic Absentee Voting Initiatives for Military and Overseas Citizens. [hyperlink, http://www.gao.gov/products/GAO-07-774]. Washington, D.C.: June 14, 2007. 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. Elections: Absentee Voting Assistance to Military and Overseas Citizens Increased for the 2004 General Election, but Challenges Remain. [hyperlink, http://www.gao.gov/products/GAO-06-521]. Washington, D.C.: April 7, 2006. Elections: Views of Selected Local Election Officials on Managing Voter Registration and Ensuring Eligible Citizens Can Vote. [hyperlink, http://www.gao.gov/products/GAO-05-997]. Washington, D.C.: September 27, 2005. Elections: Additional Data Could Help State and Local Elections Officials Maintain Accurate Voter Registration Lists. [hyperlink, http://www.gao.gov/products/GAO-05-478]. Washington, D.C.: June 10, 2005. Elections: Perspectives on Activities and Challenges Across the Nation. [hyperlink, http://www.gao.gov/products/GAO-02-3]. Washington, D.C.: October 15, 2001. Voters With Disabilities: Access to Polling Places and Alternative Voting Methods. [hyperlink, http://www.gao.gov/products/GAO-02-107]. Washington, D.C.: October 15, 2001. Elections: Voting Assistance to Military and Overseas Citizens Should Be Improved. [hyperlink, http://www.gao.gov/products/GAO-01-1026]. Washington, D.C.: September 28, 2001. Elections: The Scope of Congressional Authority in Election Administration. [hyperlink, http://www.gao.gov/products/GAO-01-470]. Washington, D.C.: March 13, 2001. [End of section] Footnotes: [1] Voter turnout varies, and in general is lower in midterm election years than in presidential election years. These turnout rates reflect total turnout (or total ballots counted) divided by the voting- eligible population. U.S. Elections Project. [2] More specifically, the 1845 federal law first established a uniform date--the Tuesday next after the first Monday in the month of November of the year in which they are to be appointed--for the appointment of Presidential and Vice Presidential electors. (January 23, 1845, ch. 1, 5 Stat. 721. The current provision is codified at 3 U.S.C. § 1.) [3] In 1872, the same (Tuesday after the first Monday in November) date was established as the day for the election of Representatives to the Congress. (February 2, 1872, ch. 11, § 3, 17 Stat. 28. The current provision is codified at 2 U.S.C. § 7.) After ratification of the Seventeenth Amendment establishing the direct popular election of Senators, a federal law in 1914 established the same date for the election of Senators to the Congress. (June 4, 1914, ch.103, § 1, 38 Stat. 384. The current provision is codified at 2 U.S.C. § 1.) [4] See, for example, H.R. 254, 111TH Cong. (2009); S. 149, 111TH Cong. (2009); H.R. 7034, 110TH Cong. (2008); H.R. 6240, 110TH Cong. (2008); S. 2638, 110TH Cong. (2008); S. 144, 109TH Cong. (2005); S. 1320, 107TH Cong. (2001); S. 1463, 105TH Cong. (1997); and H.R. 1367, 104TH Cong. (1995). [5] H.R. Rep. No. 111-202, at 60 (2009). [6] 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). See the Related GAO Products list for reports issued on elections since 2001. [7] The EAC is an independent, bipartisan commission created by the Help America Vote Act of 2002 (HAVA) to assist state and local election officials with the administration of federal elections. See U.S. Election Assistance Commission, Alternative Voting Methods (Washington, D.C.: September 2008) and U.S. Election Assistance Commission, 2010 Statutory Overview (Washington, D.C.: May 2011). [8] Center for Democracy and Election Management, Election Administration Profiles of All Fifty States (Washington, D.C.: July 15, 2009). [9] The 9 states are California, Delaware, Illinois, Louisiana, Maryland, New Hampshire, North Carolina, Texas, and Wisconsin. [10] Throughout this report, we use the term "voting by mail" to refer to a process where all votes are cast by mail. [11] We previously reported the results of web-based survey we conducted in 2005 of the 50 states and the District regarding the November 2004 general election. See GAO, Elections: 2005 Survey of State Election Officials, [hyperlink, http://www.gao.gov/products/GAO-06-451SP] (Washington, D.C.: June 6, 2006). [12] Why Tuesday? is a nonpartisan, nonprofit 501(c)(3) organization founded in 2005 to find solutions to increase voter turnout and participation in elections. [13] We selected 1985 because alternative voting methods, such as no- excuse absentee voting and in-person early voting, were generally not available prior to this time. Thus, applicable research on their effect on voter turnout was not widely conducted. [14] Specifically, we conducted an initial search of political science journals and conference proceedings for research on alternative voting methods. Senior social scientists with training and experience in evaluation research methods or political science subsequently assessed the studies found to determine whether the methods used met generally accepted social science standards. We then used a data collection instrument to systematically collect information from the 24 identified studies about their methods and conclusions. Each study was read and coded by a senior social scientist and a second senior social scientist then reviewed each completed data collection instrument to verify the accuracy of the information included. See the bibliography of this report for a listing of the studies we reviewed. [15] As described by the Supreme Court, "the States have evolved comprehensive, and in many respects, complex, election codes regulating in most substantial ways, with respect to both federal and state elections, the time, place, and manner of holding primary and general elections, the registration of voters, and the selection and qualification of candidates." Storer v. Brown, 415 U.S. 724, 730 (1974). [16] 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 second group contains 9 states that delegate election responsibility principally to subcounty governmental units. [17] See GAO, Elections: The Scope of Congressional Authority in Election Administration, [hyperlink, http://www.gao.gov/products/GAO-01-470] (Washington, D.C.: Mar. 13, 2001). [18] Article II, Section 1, Clause 4, pertains to Congress' power to set the time of choosing of presidential electors: "The Congress may determine the Time of choosing the Electors, and the Day on which they shall give their Votes; which Day shall be the same throughout the United States." Article I, Section 4, Clause 1, known as the Elections Clause, provides Congress with broad authority to regulate congressional elections: "The Times, Places, and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of choosing Senators." [19] 2 U.S.C. §§ 1, 7. [20] 3 U.S.C. § 1. [21] U.S. Election Assistance Commission, Alternative Voting Methods. [22] Jurisdictions call their poll workers by different titles, including clerks, wardens, election judges, inspectors, captains, and precinct officers. [23] [hyperlink, http://www.gao.gov/products/GAO-06-450]. [24] An optical scan voting system is composed of computer-readable ballots, appropriate marking devices, privacy booths, and a computerized tabulation machine. Voters record their choices using an appropriate writing instrument to fill in boxes or ovals or to complete an arrow next to the candidate's name or the issue. [25] Voters make their selection on DREs by pressing a button or touching the screen next to the candidate's name or ballot issue. When they are finished, they cast their votes by pressing a final "vote" button on the machine or screen. [26] Examples of excuses a voter may provide for not voting on Election Day include being sick, having a disability, being out of the country, or having religious commitments. [27] Known as the language minority provisions of the Voting Rights Act of 1965, sections 203 and 4(f)(4) of the act are designed to help members of applicable language minority groups to participate effectively in the electoral process. In general, under these provisions, covered jurisdictions must provide bilingual election materials and other assistance to protect the voting rights of U.S. citizens of certain ethnic groups whose command of the English language may be limited. For more on this subject, see GAO, Bilingual Voting Assistance: Selected Jurisdictions' Strategies for Identifying Needs and Providing Assistance, [hyperlink, http://www.gao.gov/products/GAO-08-182] (Washington, D.C.: Jan. 18, 2008). [28] The Voting Accessibility for the Elderly and Handicapped Act requires that, with a few exceptions, political subdivisions responsible for conducting elections ensure that polling places used in federal elections are accessible to voters with disabilities. HAVA established additional requirements. [29] In addition, other states might have had provisions that allowed certain counties or precincts to administer elections entirely by mail. For example, in California--one of nine states in which we conducted interviews--state law authorizes election officials to designate precincts with 250 or fewer registered voters as "mail ballot precincts," where citizens who want to vote can only do so by mail. In California's two least populous counties, all of the precincts contain fewer than 250 registered voters and, thus, all elections in these counties are conducted entirely by mail. All three of the local jurisdictions where we conducted interviews in California had some precincts that were mail ballot precincts in the November 2010 general election. [30] These eight states were Arizona, California, Colorado, Hawaii, Montana, New Jersey, Utah, and Washington. [31] See [hyperlink, http://www.gao.gov/products/GAO-06-450]. This information was based on the result of web-based surveys we conducted in 2005 of the 50 states and the District. See [hyperlink, http://www.gao.gov/products/GAO-06-451SP] for additional survey results. [32] According to state election officials we interviewed from both of these states, there has been minimal interest by either state's legislature in considering implementation of these alternative voting methods. According to a senior Delaware official, the state would face challenges conducting early voting or handling an increase in no- excuse absentee ballots because of limitations in their current voting technology. According to a senior New Hampshire official, the state's culture supports the idea that election intensity (i.e., limiting voting options to Election Day) produces higher voter participation than other alternatives. [33] Maryland early voting statutory provisions require the inclusion of Sunday hours between 12:00 p.m. and 6:00 p.m. in the early voting period for the 2012 presidential primary and general election. [34] For example, in counties with populations of 100,000 or more, Texas early voting statutory provisions outline the minimum number of hours that early voting must be provided on both the last Saturday and Sunday of the early voting period. [35] In this section of the report, we discuss together responses from election officials we interviewed in the District and local election jurisdictions because the District's responsibilities for conducting federal elections are generally similar to those of local jurisdictions. [36] In 2006 we reported that, based on a nationwide survey of local election jurisdictions we conducted in 2005, an estimated 89 percent of jurisdictions relied in part on lists or rosters of poll workers from past elections to recruit poll workers for the November 2004 general election. For more information, see GAO-06-450. [37] Election officials from the other 14 local jurisdictions we interviewed did not express views or provide information specifically on how moving the date of federal elections might affect their ability to recruit from the poll workers and sources they have relied on in the past. Although we asked election officials in nine states, the District, and 17 local jurisdictions about whether or not various aspects of the election process might be affected by changing Election Day to a weekend, not all expressed views or provided information on every specific issue discussed throughout this report. [38] According to EAC, the following requirements apply to all polling places: (1) buildings should be located in close proximity to the voters who are served by the location; (2) property owners must grant permission for the use of their building for voting purposes on Election Day; (3) the buildings selected for use as polling places must meet federal and state accessibility requirements; (4) buildings must have an adequate-sized room or hallway sufficient to meet the needs for setting up equipment and voter check-in stations, including adequate space for voters to wait in line; (5) property owners may be requested to open the building for poll worker use the evening prior to Election Day, and must be willing to open the building in the early morning hours of Election Day until after the close of the polls; (6) the property must have sufficient parking available for voters' use; and (7) traffic ingress/egress must be evaluated, based on the number of expected voters on Election Day. [39] [hyperlink, http://www.gao.gov/products/GAO-06-450]. [40] [hyperlink, http://www.gao.gov/products/GAO-06-450]. [41] Election officials arrived at this estimate based on hiring two private security guards for each of their 183 polling places for 9 hours (at a rate of $30 per hour) over the Saturday night of an election weekend, when they would expect polls would be closed. [42] Under the provisions of H.R. 254, for presidential and congressional general elections, polling places would have been required to be open beginning on Saturday at 10:00 a.m. eastern standard time and ending on Sunday at 6:00 p.m. eastern standard time. H.R. 254 would have additionally authorized a polling place to close between the hours of 10:00 p.m. local time on Saturday and 6:00 a.m. local time on Sunday as provided by the law of the state in which the polling place is located. [43] In the November 2, 2010, Consolidated General Election, San Francisco voters adopted a ballot measure creating a Saturday Voting Pilot Program that required the Department of Elections to open voting locations for each precinct on Saturday, November 5, 2011, prior to Election Day on Tuesday, November 8, 2011, if sufficient funds to cover all Saturday costs were raised from private donations. Since no contributions were received by the set deadline, the Saturday portion of the election was canceled. The department had estimated the cost of the 2-day election would be about $7.5 million, with the cost associated with opening all polling places on Saturday to be about $2.4 million. [44] EAC, among other things, is responsible for testing, certifying, decertifying, and recertifying voting system hardware and software. In 2007, EAC established and began implementing its voting system certification program. State participation in this certification program is optional. According to EAC, some states, through legislation or administrative rules, mandate participation to varying degrees. EAC certification means that a voting system has been successfully tested by an accredited voting system test laboratory, meets requirements set forth in a specific set of federal voting system standards, and performs according to the manufacturer's specifications. A previously certified voting system must be recertified when a modification is made that would require further testing and review by the EAC. [45] Officials from one jurisdiction with two election staff said that while a weekend election would increase their workload, their associated costs would not increase because both staff are elected positions and therefore not eligible for overtime pay. [46] One could make predictions about the possible effect by making assumptions about whether weekend elections would make voting more convenient, but the accuracy of these assumptions could not be directly verified using evidence from actual federal elections. [47] According to a 2007 study commissioned by the Canadian government, at least 20 countries, including Brazil, held federal elections on Sunday and at least 3 countries, including Australia, held federal elections on Saturday. See "Potential Impacts of Extended Advance Voting on Voter Turnout," a report prepared at the request of Elections Canada (September 2007). [48] These two countries, among others, are often referred to as having "compulsory voting." In Australia, for example, under the Commonwealth Electoral Act 1918, an elector is required to, among other things, attend a voting place and receive and cast a ballot. [49] The studies we reviewed are listed in the bibliography to this report. [50] The variety of methodologies used, in addition to chance variation, helps explain why the estimated effects varied across studies. The 24 studies used varied elections, jurisdictions, and time periods to evaluate the effect of alternative voting methods. Researchers analyzed turnout in elections for President, Congress, and state and local offices held in various periods from 1960 through 2008. Some researchers studied many elections across all 50 states, while others studied many voters or precincts in a smaller number of states or shorter time periods. [51] For reviews of studies making similar conclusions, see Paul Gronke, et al., "Convenience Voting," Annual Review of Political Science (2008), 437-455, and "Potential Impacts of Extended Advance Voting on Voter Turnout," a report prepared at the request of Elections Canada, 25. [52] See Priscilla L. Southwell and Justin I. Burchett, "The Effect of All-Mail Elections on Voter Turnout," American Politics Quarterly, vol. 28 (2000), 72-79, and Paul Gronke and Peter A.M. Miller, "Voting by Mail and Turnout: A Replication and Extension," paper presented at the 2007 Annual Meeting of the American Political Science Association, Chicago, Ill. [53] See, for example, Raymond E. Wolfinger and Stephen J. Rosenstone, Who Votes? (New Haven, Conn.: Yale University Press, 1980); Steven J. Rosenstone and John Mark Hansen, Mobilization, Participation, and Democracy in America (New York, N.Y.: Macmillan Publishing, 1993); and Sidney Verba, Kay Lehman Schlozman, and Henry E. Brady, Voice and Equality: Civic Voluntarism in American Politics (Cambridge, Mass.: Harvard University Press, 1995). [54] Rosenstone and Hansen, Mobilization, 273-275. [55] Raymond E. Wolfinger, Benjamin Highton, and Megan Mullin, "How Postregistration Laws Affect the Turnout of Citizens Registered to Vote," State Politics and Policy Quarterly, vol. 5 (2005), 1-23. [56] Adam J. Berinsky, Nancy Burns, and Michael W. Traugott, "Who Votes By Mail? A Dynamic Model of the Individual-Level Consequences of Voting-By-Mail Systems," The Public Opinion Quarterly, vol. 65 (2001), 178-197. [57] The turnout statistics discussed here may not match certified vote totals for Maryland jurisdictions because we analyzed a custom database of citizens registered for the 2010 general election. [58] These comparisons do not account for the fact that Maryland offered 5 days of weekday voting but only 1 day of weekend voting in 2010. See app. II for a more-detailed analysis of how the turnout varied through the voting period. [59] Known as the language minority provisions of the Voting Rights Act, sections 203 and 4(f)(4) of the act are designed to help members of applicable language minority groups to participate effectively in the electoral process. In general, under these provisions, covered jurisdictions must provide bilingual election materials and other assistance to protect the voting rights of U.S. citizens of certain ethnic groups whose command of the English language may be limited. [60] Section 10.301.1(b) of the Maryland Election Law Code specifies the number of early voting centers for each county. If a county has less than 150,000 registered voters, the county shall have one early voting center. If a county has more than 150,000 but fewer than 300,000 registered voters, the county shall have three early voting centers. If a county has more than 300,000 registered voters, the county shall have five early voting centers. [61] The turnout statistics discussed here may not match certified vote totals for Maryland jurisdictions because we analyzed a custom database of citizens registered for the 2010 general election. [62] We rescaled calendar time to poll days to avoid analyzing periods when the polls were closed. In effect, this adjusts the voting duration times for the time "at risk" of voting. While Maryland standardized early voting poll hours across counties, we included voting times outside of the official poll hours, which may have represented citizens who were in line to vote when the polls closed. As a result, we defined the start and end of each poll day as the earliest and latest recorded voting time on a particular calendar day of early voting. [63] The data represent the empirical cumulative distribution function of the voting times. This implies that the first derivative of the curve at voting time t equals the density, which can be interpreted as the instantaneous turnout rate at t or the probability of voting within an arbitrarily small neighborhood of t. [64] Instead of using hierarchical and multinomial models of the turnout and voting method decisions, we used simpler models with flexible response functions to simply quantify descriptive patterns in the raw data, in part due to the limited number of covariates available. For similar reasons, we did not include prior turnout as a covariate to avoid complications associated with lagged response variables. [65] See 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). This information was based on the result of web-based surveys we conducted in 2005 of the 50 states and the District. See GAO, Elections: 2005 Survey of State Election Officials, [hyperlink, http://www.gao.gov/products/GAO-06-451SP] (Washington, D.C.: June 6, 2006) for additional survey results. [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: Ralph Dawn, Managing Director, dawnr@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.