Center for Democracy and Election Management
Carnegie Corporation of New York
The Ford Foundation
John S. and James L. Knight Foundation
Electionline.org/The Pew Charitable Trusts
Letter from the Co-Chairs
Preface by the Executive Director
1: Goals and Challenges of Election Reform
1.1 Help America Vote Act: Strengths and Limitations
1.2 Learning from the World
1.3 Transforming the Electoral System – Five Pillars
1.4 Urgency of Reform
2: Voter Registration and Identification
2.1 Uniformity Within States – Top-Down Registration Systems
2.2 Interoperability Among States
2.3 Provisional Ballots
2.4 Communicating Registration Information
2.5 Voter Identification
2.6 Quality in Voter Registration Lists
3: Voting Technology
3.1 Voting Machines
3.3 Security for Voting Systems
3.4 Internet Voting
4: Expanding Access to Elections
4.1 Assured Access to Elections
4.2 Vote by Mail
4.3 Vote Centers
4.4 Military and Overseas Voting
4.5 Access for Voters with Disabilities
4.6 Re-Enfranchisement of Ex-Felons
4.7 Voter and Civic Education
5: Improving Ballot Integrity
5.1 Investigation and Prosecution of Election Fraud
5.2 Absentee Ballot and Voter Registration Fraud
6: Election Administration
6.2 Poll Worker Recruitment
6.3 Polling Station Operations
6.4 Research on Election Management
6.5 Cost of Elections
7: Responsible Media Coverage
7.1 Media Access for Candidates
7.2 Media Projections of Election Results
8: Election Observation
9: Presidential Primary and Post-Election Schedules
9.1 Presidential Primary Schedule
9.2 Post-Election Timeline
Appendix: Estimated Costs of Recommended Improvements
Summary of Recommendations
About the Commission on Federal Election Reform
Elections are the heart of democracy. They are the instrument for the people to choose leaders and hold them accountable. At the same time, elections are a core public function upon which all other government responsibilities depend. If elections are defective, the entire democratic system is at risk.
Americans are losing confidence in the fairness of elections, and while we do not face a crisis today, we need to address the problems of our electoral system.
Our Commission on Federal Election Reform was formed to recommend ways to raise confidence in the electoral system. Many Americans thought that one report — the Carter-Ford Commission — and one law — the Help America Vote Act of 2002 (HAVA) — would be enough to fix the system. It isn’t. In this report, we seek to build on the historic achievement of HAVA and put forward a bold set of proposals to modernize our electoral system.
Some Americans will prefer some of our proposals to others. Indeed, while all of the Commission members endorse the judgments and general policy thrust of the report in its entirety, they do not necessarily support every word and recommendation. Benefitting from Commission members with diverse perspectives, we have proposed, for example, a formula for transcending the sterile debate between integrity and access. Twenty-four states now require identification for voters, with some systems likely to restrict registration. We are recommending a photo ID system for voters designed to increase registration with a more affirmative and aggressive role for states in finding new voters and providing free IDs for those without driver’s licenses. The formula we recommend will result in both more integrity and more access. A few of our members have expressed an alternative view of this issue.
Still, our entire Commission is united in the view that electoral reform is essential and that our recommended package of proposals represents the best way to modernize our electoral system. We urge all Americans, including the legislative and executive branches of government at all levels, to recognize the urgency of election reform and to seriously consider the comprehensive approach outlined herein.
We present this report because we believe the time for acting to improve our election system is now.
Jimmy Carter James A. Baker, III
Co-Chairs of the Commission on Federal Election Reform
^ Back to top
Polls indicate that many Americans lack confidence in the electoral system, but the political parties are so divided that serious electoral reform is unlikely without a strong bipartisan voice. Our country therefore owes a great debt to former President Jimmy Carter and former Secretary of State James A. Baker, III for leading this Commission and forging a plan for election reform.
To build confidence, the Commission recommends a modern electoral system built on five pillars: (1) a universal and up-to-date registration list, accessible to the public; (2) a uniform voter identification system that is implemented in a way that increases, not impedes, participation; (3) measures to enhance ballot integrity and voter access; (4) a voter-verifiable paper trail and improved security of voting systems; and (5) electoral institutions that are impartial, professional, and independent. Democrats, Republicans, and Independents tend to prefer different elements of this package, but President Carter and Secretary Baker drew strength rather than stalemate from the diverse perspectives in fashioning an approach that is greater than the sum of these parts.
Our Commission was fortunate to have an outstanding staff and academic advisors, and we have benefited from advice by Members of Congress and staff, election officials, and representatives of a wide range of non-governmental organizations devoted to improving our democracy. See our website for a list of advisors and the studies and testimony: http://www.american.edu/Carter-Baker.
We acknowledge the support of many at the end of this report, but let me identify here a few people whose work was crucial to the Commission: Daniel Calingaert, the Associate Director of American University’s Center for Democracy and Election Management, Doug Chapin of Electionline.org, John Williams, Senior Advisor to Secretary Baker, Kay Stimson, Media Liaison, and Murray Gormly, Administrative Coordinator. The Commission was organized by American University’s Center for Democracy and Election Management. We are also grateful to the James A. Baker III Institute for Public Policy of Rice University and The Carter Center for hosting the other two meetings.
Finally, the Commission could not have accomplished its goal without the generosity of its funders and the advice and support of the following individuals: Geri Mannion of the Carnegie Corporation; Thomasina Williams of the Ford Foundation; Julie Kohler of the John S. and James L. Knight Foundation; Dena Jones of Omidyar Network, and The Pew Charitable Trusts.
At AU’s Center for Democracy and Election Management, we view this Commission as a major step toward developing the educational foundation for students, professionals, and the public to deepen our understanding of democracy and elections in the United States and the world.
Robert A. Pastor,
^ Back to top
Building confidence in U.S. elections is central to our nation’s democracy. At a time when there is growing skepticism with our electoral system, the Commission believes that a bold new approach is essential. The Commission envisions a system that makes Americans proud of themselves as citizens and of democracy in the United States. We should have an electoral system where registering to vote is convenient, voting is efficient and pleasant, voting machines work properly, fraud is deterred, and disputes are handled fairly and expeditiously.
This report represents a comprehensive proposal for modernizing our electoral system. We propose
to construct the new edifice for elections on five pillars:
First, we propose a universal voter registration system in which the states, not local jurisdictions, are responsible for the accuracy and quality of the voter lists. Additionally, we propose that the U.S. Election Assistance Commission (EAC) develop a mechanism to connect all states’ list. These topdown and interoperable registration lists will, if implemented successfully, eliminate the vast majority of complaints currently leveled against the election system. States will retain control over their registration list, but a distributed database can remove interstate duplicates and help states to maintain an up-to-date, fully accurate registration list. This would mean people would need to register only once in their lifetime, and it would be easy to update their registration information when they move. We also propose that all states establish uniform procedures for counting provisional ballots, and many members recommend that the ballots should be counted if the citizen has voted in the correct jurisdiction.
Second, to make sure that a person arriving at a polling site is the same one who is named on the list, we propose a uniform system of voter identification based on the “REAL ID card” or an equivalent for people without a drivers license. To prevent the ID from being a barrier to voting, we recommend that states use the registration and ID process to enfranchise more voters than ever. States should play an affirmative role in reaching out to non-drivers by providing more offices, including mobile ones, to register voters and provide photo IDs free of charge. There is likely to be less discrimination against minorities if there is a single, uniform ID, than if poll workers can apply multiple standards. In addition, we suggest procedural and institutional safeguards to make sure that the rights of citizens are not abused and that voters will not be disenfranchised because of an ID requirement. We also propose that voters who do not have a photo ID during a transitional period receive a provisional ballot that would be counted if their signature is verified.
Third, we propose measures that will increase voting participation by having the states assume greater responsibility to register citizens, make voting more convenient, and offer more information on registration lists and voting. States should allow experimentation with voting centers. We propose ways to facilitate voting by overseas military and civilians and ways to make sure that people with disabilities have full access to voting. In addition, we ask the states to allow for restoration of voting rights for ex-felons (other than individuals convicted of capital crimes or registered sex offenders) when they have fully served their sentence. We also identify several voter and civic education programs that could increase participation and inform voters, for example, by providing information on candidates and the voting process to citizens before the election. States and local jurisdictions should use Web sites, toll-free numbers, and other means to inform citizens about their registration status and the location of their precinct.
To improve ballot integrity, we propose that federal, state, and local prosecutors issue public reports on their investigations of election fraud, and we recommend federal legislation to deter or prosecute systemic efforts to deceive or intimidate voters. States should not discourage legal voter registration or get-out-the-vote activities, but they need to do more to prevent voter registration and absentee ballot fraud.
Fourth, we propose ways to give confidence to voters using electronic voting machines that their votes will be counted accurately. We call for an auditable backup on paper at this time, but we recognize the possibility of alternative technologies to audit those machines in the future. We encourage independent testing of voting systems (to include voting machines and software source code) under EAC supervision.
Finally, we recommend strengthening and restructuring the system by which elections have been administered in our country. We propose that the EAC and state election management bodies be reconstituted on a nonpartisan basis to become more independent and effective. We cannot build confidence in elections if secretaries of state responsible for certifying votes are simultaneously chairing political campaigns, and the EAC cannot undertake the additional responsibilities recommended by this report, including critical research, without gaining additional funds and support. Polling stations should be organized to reduce the chances of long lines; they should maintain “log-books” on Election Day to record complaints; and they need electronic poll-books to help voters find their correct precinct. HAVA should be fully funded and implemented by 2006.
The Commission puts forward 87 specific recommendations. Here are a few of the others:
^ Back to top
The vigor of American democracy rests on the vote of each citizen. Only when citizens can freely and privately exercise their right to vote and have their vote recorded correctly can they hold their leaders accountable. Democracy is endangered when people believe that their votes do not matter or are not counted correctly.
Much has happened since November 2000, when many Americans first recognized that their electoral system had serious problems with flawed voter registration lists, obsolete voting machines, poorly designed ballots, and inadequate procedures for interpreting disputed votes. Congress and the President, Democrats and Republicans, responded with a truly historic initiative – the Help America Vote Act of 2002 (HAVA), the first comprehensive federal law in our nation’s history on electoral administration. The law represents a significant step forward, but it falls short of fully modernizing our electoral system.
On the eve of the November 2004 election, a New York Times poll reported that only onethird of the American people said that they had a lot of confidence that their votes would be counted properly, and 29 percent said they were very or somewhat concerned that they would encounter problems at the polls. Aware of this unease, the U.S. Department of Justice deployed 1,090 election observers — more than three times the number sent in 2000.1 After the election, a minority of Americans — only 48 percent — said they were very confident that the votes cast across the country were accurately counted, according to a Pew Research Center survey. Thirty-seven percent had doubts (somewhat confident), and 14 percent were not confident that the votes were accurately counted.2With a strong desire to contribute to building confidence in our electoral process, this Commission came together to analyze the state of the electoral system, to assess HAVA’s implementation, and to offer recommendations for further improvement. Public confidence in the electoral system is critical for our nation’s democracy. Little can undermine democracy more than a widespread belief among the people that elections are neither fair nor legitimate. We believe that further important improvements are necessary to remove any doubts about the electoral process and to help Americans look upon the process of casting their ballot as an inspiring experience — not an ordeal.
We address this report to the American people and to the President, Congress, U.S. Election Assistance Commission, states, election administrators, and the media. Our recommendations aim both to increase voter participation and to assure the integrity of the electoral system. To achieve those goals, we need an accurate list of registered voters, adequate voter identification, voting technology that precisely records and tabulates votes and is subject to verification, and capable, fair, and nonpartisan election administration.
While each state will retain fundamental control over its electoral system, the federal government should seek to ensure that all qualified voters have an equal opportunity to exercise their right to vote. This will require greater uniformity of some voting requirements and registration lists that are accurate and compatible among states. Greater uniformity is also needed within states on some voting rules and procedures. The federal government should fund research and development of voting technology that will make the counting of votes more transparent, accurate, and verifiable.
The Help America Vote Act of 2002 (HAVA) established numerous federal requirements for state and local election administration in exchange for a promise of $3.97 billion in federal funding, of which approximately $3.1 billion has been appropriated to date. These requirements reflected a national consensus on the general outline of reform, best represented by the 2001 report of the National Commission on Federal Election Reform, co-chaired by former Presidents Jimmy Carter and Gerald Ford. HAVA’s mandates were adopted as part of a compromise between the parties on the divisive issue of access to the ballot (largely championed by Democrats and their allies) versus protecting the integrity of the electoral process (generally favored by Republicans and their supporters).
Under this compromise, described by its sponsors as making it “easier to vote and harder to cheat,” HAVA sought to lower barriers to voting while establishing somewhat tighter controls on registration and voter identification. Consequently, HAVA’s mandates focused on four major requirements: (1) statewide computerized voter lists; (2) voter ID for individuals who register by mail but do not provide it when registering; (3) provisional ballots for voters whose names are missing from the registration rolls on Election Day; and (4) measures to make voting more accessible for voters with disabilities. The main provisions of HAVA are as follows:
Under HAVA, states are required to complete their statewide voter databases by January 1, 2006, and some expenditures of HAVA funds will extend well beyond that date. Our Commission therefore calls for full implementation and full funding of HAVA.
The first presidential election after HAVA became law — on November 2, 2004 — brought to light as many problems as in 2000, if not more. HAVA, which will take years to be fully implemented, was not responsible for most of the complaints. Instead, voters were discouraged or prevented from voting by the failure of election offices to process voter registration applications or to mail absentee ballots in time, and by the poor service and long lines at polling stations in a number of states. There were also reports of improper requests for voter ID and of voter intimidation and suppression tactics. Concerns were raised about partisan purges of voter registration lists and about deliberate failures to deliver voter registration applications to election authorities. Moreover, computer malfunctions impugned election results for at least one race, and different procedures for counting provisional ballots within and between states led to legal challenges and political protests. Had the margin of victory for the presidential contest been narrower, the lengthy dispute that followed the 2000 election could have been repeated.
The November 2004 elections also showed that irregularities and fraud still occur. In Washington, for example, where Christine Gregoire was elected governor by a 129-vote margin, the elections superintendent of King County testified during a subsequent unsuccessful election challenge that ineligible ex-felons had voted and that votes had been cast in the names of the dead. However, the judge accepted Gregoire’s victory because with the exception of four ex-felons who admitted to voting for Dino Rossi, the authorities could not determine for whom the other illegal votes were cast. In Milwaukee, Wisconsin, investigators said they found clear evidence of fraud, including more than 200 cases of felons voting illegally and more than 100 people who voted twice, used fake names or false addresses, or voted in the name of a dead person. Moreover, there were 4,500 more votes cast than voters listed.3 One potential source of election fraud arises from inactive or ineligible voters left on voter registration lists. By one estimate, for example, there were over 181,000 dead people listed on the voter rolls in six swing states in the November 2004 elections, including almost 65,000 dead people listed on the voter rolls in Florida.4
Some of these problems may be addressed by the full implementation of HAVA, but it is clear that others will not. Due to vague mandates on provisional voting and identification cards, counties and states applied different standards. This led to a significant proliferation of legal challenges. A closer presidential election likely would have brought an avalanche of litigation. HAVA does not address interoperable registration lists among states, and it is also vague as to whether states should create a top-down, statecontrolled registration list or a bottom-up list controlled by local election administrators. The weak structure of the U.S. Election Assistance Commission, a product of a HAVA compromise, has stymied its ability to be clear or authoritative on almost any subject, even on whether to verify electronic machine votes with paper ballots. Thus, there is a compelling need for further election reform that builds on HAVA.
One of the most important laws on the right of Americans to vote is the Voting Rights Act of 1965. Key provisions of the Act are due to expire in 2007. These include the language provision (Section 203), which requires jurisdictions to provide voting materials in minority languages in areas where language minority groups make up a significant portion of the population, and the pre-clearance provision (Section 5), which requires federal pre-clearance for all changes to voting rules or procedures made by specified jurisdictions with a history of voter discrimination. Our Commission believes this Act is of the utmost importance.
1.1.1 The Help America Vote Act should be fully implemented by 2006, as mandated by the law, and fully funded.
1.1.2 The Commission urges that the Voting Rights Act be vigorously enforced and that Congress and the President seriously consider reauthorizing those provisions of the Act that are due to expire in 2007.
In its deliberations, our Commission considered the best practices of election systems around the world. Many other democracies achieve significantly higher levels of voter participation due, in part, to more effective voter registration. Election authorities take the initiative to contact and register voters and conduct audits of voter registration lists to assure that they are accurate. In addition, voter registration in many countries is often tied directly to a voter ID, so that voter identification can enhance ballot integrity without raising barriers to voting. Voters in nearly 100 democracies use a photo identification card without fear of infringement on their rights.5
Nonpartisan election administration has also proved effective abroad. Over the past three decades, election management institutions have evolved in many other democracies. Governments had previously conducted elections, but as concern was raised that they might give advantage to incumbents, independent election commissions were formed. Initially, election commissioners in other countries frequently represented political parties, but they often stalemated or reached agreement with each other at the public’s expense. This explains why the trend in the world is toward independent election commissions composed of nonpartisan officials, who serve like judges, independently of the executive or legislative branches (see Table 5 on page 52). Political party representatives can observe deliberations on these commissions but not vote on decisions. Nonpartisan election officials are generally regarded as fair arbiters of the electoral process who make their best efforts to administer elections impartially and effectively.
^ Back to top
The recommendations of our Commission on Federal Election Reform aim both to increase voter participation and to assure the integrity of the electoral system. To accomplish these goals, the electoral system we envision should be constructed on the following five sturdy pillars:
An electoral system built on these pillars will give confidence to all citizens and will contribute to high voter participation. The electoral system should also be designed to reduce the possibility or opportunity for litigation before, and especially after, an election. Citizens should be confident that the results of the election reflect their decision, not a litigated outcome determined by lawyers and judges. This is achieved by clear and unambiguous rules for the conduct of the election established well in advance of Election Day.
The ultimate test of an election system is its ability to withstand intense public scrutiny during a very close election. Several close elections have taken place in recent years, and our election system has not always passed that test. We need a better election system.
Although the public continues to call for election reform, and several election bills have been introduced, the issue is low on the Congress’s agenda at this time. Some congressional leaders believe that further reform should wait until HAVA is fully implemented. We believe that the need for additional electoral reform is abundantly clear, and our recommendations will bolster HAVA to further strengthen public confidence in the electoral process. If we wait until late 2006, we will lose the opportunity to put new reforms in place for the 2008 elections, and as a result, the next presidential election could be fraught with problems. Electoral reform may stay out of public view until the 2006 elections begin to approach, but by that time, it may be too late. We need Congress to press ahead with election reform now. Indeed, election reform is best accomplished when it is undertaken before the passions of a specific election cycle begin.
We are Republicans, Democrats, and Independents. But we have deliberately attempted to address electoral issues without asking the question as to whether a particular political party would benefit from a particular reform. We have done so because our country needs a clear unified voice calling for serious election reform. Congress has been reluctant to undertake reform, in part because members fear it could affect their chances of re-election and, when finally pressed by the public, Democrats and Republicans have addressed each reform by first asking whether it would help or harm each party’s political prospects. This has proven to be not only a shortsighted but also a mistaken approach. Despite widespread belief that two recent reforms — the National Voter Registration Act of 1993 and the Bipartisan Campaign Finance Reform of 2002 — would advantage Democrats at the expense of Republicans, evidence suggests such beliefs were wrong. Having a fair electoral process in which all eligible citizens have an opportunity to participate freely is a goal that transcends any individual partisan interest. This assures the winning candidates the authority to legitimately assume office. For the losing candidate it assures that the decision can be accepted as the will of the voters.
Our recommendations are aimed at several timeframes and audiences. Some require immediate action, and others can be considered later. We propose some for the federal government and some for the states. But we have offered all the recommendations based on our views as to how they can best help our country — not our political parties. Together, these reforms should catalyze a shift in the way that elections are administered. We hope they will not only restore American confidence in our elections, but also strengthen the respect from those in the world who look to our democracy as a model.
^ Back to top
Effective voter registration and voter identification are bedrocks of a modern election system. By assuring uniformity to both voter registration and voter identification, and by having states play an active role in registering as many qualified citizens as possible, access to elections and ballot integrity will both be enhanced. These steps could help bring to an end the sterile debate between Democrats and Republicans on access versus integrity.
The most common problems on Election Day concern voter registration (see Table 1 on page 17). Voter registration lists often are riddled with inaccuracies because Americans are highly mobile, and local authorities, who have maintained most lists, are poorly positioned to add and delete names of voters who move within or between states. To comprehend the magnitude of this challenge, consider the following. During the last decade, on average, about 41.5 million Americans moved each year. Of those, about 31.2 million moved within the same state, and 8.9 million moved to a different state or abroad. Young Americans (aged 20 to 29), representing 14 percent of the U.S. population, moved to a different state at almost three times the rate of the rest of the population.6 The process of registering voters should be made easier, and renewal due to a change of address should be made still easier.
In response to the challenge of building and maintaining better registration lists, HAVA requires states to establish statewide, computer-based registration lists that are interactive within each state by January 1, 2006. HAVA also requires provisional ballots for eligible voters who seek to vote within their jurisdiction but who are denied a ballot because their name is not found on the voter roll or because they are otherwise challenged by an election official as being ineligible to vote.
Although few states have completed their new statewide voter databases, the limitations of the existing efforts are already clear. Several states have left the primary responsibility for voter lists in the hands of counties and municipalities. There is little if any effort to assure quality in statewide voter databases. The U.S. Election Assistance Commission (EAC) has not assessed the quality of statewide voter databases and is unlikely to do so in the future. Moreover, it has provided only vague guidance to states on how to organize their voter registration lists — on even the most basic question as to whether states or counties should be in charge.
In addition to statewide registration systems and provisional ballots, HAVA requires that states insist on voter identification only when a person has registered by mail for the first time in a federal election. This provision, like the others, was implemented very differently across the country, with some areas not even applying the minimum requirement. Since HAVA, an increasing number of states have insisted on stringent, though very different, ID requirements for all voters. This, in turn, has caused concern that such requirements could erect a new barrier to voting for people who do not have the requisite identification card. Georgia, for example, introduced a new law in July 2005 that requires all voters to show a government-issued photo ID at the polls.Although there are 159 counties, only 56 locations in the entire state issue such IDs, and citizens must either pay a fee for the ID or declare indigence.
While states will retain principal responsibility for the conduct of elections, greater uniformity in procedures for voter registration and identification is essential to guarantee the free exercise of the vote by all U.S. citizens. The EAC should facilitate greater uniformity in voter registration and identification procedures and should be empowered to do so by granting and withholding federal funds to the states. If Congress does not appropriate the funds, then we recommend that it amend the law to require uniformity of standards.
A complete, accurate, and current voter roll is essential to ensure that every eligible citizen who wants to vote can do so, that individuals who are ineligible cannot vote, and that citizens cannot vote more than once in the same election. A voter registration list must contain all eligible voters (including new registrants) and must contain correct information concerning the voter’s identity and residence.
Incomplete or inaccurate registration lists lie at the root of most problems encountered in U.S. elections. When a voter list omits the names of citizens who believe they properly registered or contains incorrect or out-of-date information on registered voters, eligible citizens often are denied the right to vote. Invalid voter files, which contain ineligible, duplicate, fictional, or deceased voters, are an invitation to fraud.
One reason for flawed lists is decentralized management. Local authorities often fail to delete the names of voters who move from one jurisdiction to another, and thus the lists are often inflated. For this reason, the Carter-Ford National Commission on Federal Election Reform recommended the creation of statewide voter registration systems, and this recommendation was codified into law in HAVA.
HAVA requires each state to create a “single, uniform, official, centralized, interactive computerized statewide voter registration list defined, maintained, and administered at the state level.” But states have not carried out this requirement in a consistent manner. Some are creating a “top-down” voter registration system, in which local election authorities supply information to a unified database maintained by the state. Others rely on a “bottom-up” system, whereby counties and municipalities retain their own registration lists and submit information to a state compilation of local databases at regular intervals. Top-down databases typically deliver information in real time — counties can see changes from other localities as these changes are made to the voter list. Bottom-up systems may continuethe problems that gave rise to flawed registration lists — i.e., counties retain control of the lists. Counties might not delete the names of voters who move or might not add the names of voters who register at motor vehicle bureaus or other state agencies under the National Voter Registration Act (NVRA or “Motor Voter”). Thus, the statewide lists might be different from the controlling county lists. Having two inconsistent voter lists is like a person with two watches who never knows what time it is. It is essential to have a single, accurate, current voter list.
As of June 2005, 38 states were establishing topdown voter registration systems. The remaining states were either (a) building bottom-up systems; or (b) creating systems with both top-down and bottom-up elements. Three states had not finalized plans.7 The EAC, in its interpretation of the HAVA requirement on statewide voter databases, expressed a preference for top-down systems for voter registration but did not insist on it and did not rule out bottom-up systems.
In the judgment of our Commission, bottom-up systems are not capable of providing a complete, accurate, current, and valid voter registration list. They are ineffective in removing duplicate registrations of individuals who move from one county to another and in coordinating with databases of other state agencies. Even in the best of circumstances, with excellent cooperation and interaction between states and counties — an unlikely scenario with the bottom-up system — there will be a time lag in updating voter files in a bottom-up system. This time lag could be particularly harmful in the period approaching the deadline for voters to register.
2.1.1 The Commission recommends that states be required to establish unified, top-down voter registration systems, whereby the state election office has clear authority to register voters and maintain the registration list. Counties and municipalities should assist the state with voter registration, rather than have the state assist the localities. Moreover, Congress should appropriate funds for disbursement by the U.S. Election Assistance Commission (EAC) to states to complete top-down voter registration systems.
Interoperable state voter databases are needed to facilitate updates in the registration of voters who move to another state and to eliminate duplicate registrations, which are a source of potential fraud. Approximately 9 million people move to another state or abroad each year, or about one in eight Americans between each presidential election. Such interoperability is possible because state voter databases that are centralized can be made to communicate with each other.
The limited information available on duplicate registrations indicates that a substantial number of Americans are registered to vote in two different states. According to news reports, Florida has more than 140,000 voters who apparently are registered in four other states (in Georgia, Ohio, New York, and North Carolina).8 This includes almost 46,000 voters from New York City alone who are registered to vote in Florida as well. Voting records of the 2000 elections appear to indicate that more than 2,000 people voted in two states. Duplicate registrations are also seen elsewhere. As many as 60,000 voters are reportedly registered in both North Carolina and South Carolina.9
Current procedures for updating the registration of voters who move to another state are weak or nonexistent. When people register to vote, they are usually asked to provide their prior address, so that the jurisdiction where they lived can be notified to delete their names from the voter list. Such notification, however, often does not occur. When a voter moves from Virginia to Illinois, for example, a four-step process is required to update voter registration: (1) election authorities in Illinois must ask for prior address; (2) the voter must provide prior address; (3) Illinois election authorities must notify the correct election authorities in Virginia; and (4) Virginia election authorities must remove the voter from its list. Unless all four steps are taken, this voter will remain on the voter list in Virginia. In fact, states often fail to share data or notify each other of voters who move. As a result, a substantial number of Americans are registered to vote in more than one state.
Duplicate registrations have accumulated over the years not just because there are no systems to remove them other than the one described above, but also because people who own homes in two states can register to vote in both places. In fact, when 1,700 voters who were registered in both New York and Florida requested absentee ballots to be mailed to their home in the other state, no one ever bothered to investigate.10
Interoperability among state voter databases is needed to identify and remove duplicate registrations of citizens who are registered to vote in more than one state. To make the state voter databases interoperable, the Commission recommends the introduction of a uniform template, shared voter data, and a system to transfer voter data across states.11
The template will define a common set of voter data that all states will collect in their voter databases and will share with each other. This set of data will consist of each person’s full legal name, date and place of birth, signature captured as a digital image, and Social Security number. The signature is needed to confirm the identity of voters who vote by mail.
Under HAVA, voter databases need a “unique identifier,” which is a number used to distinguish each individual, particularly those with the same or similar names. Some states use the driver’s license number as the unique identifier for voter registration. In other states, the unique identifier is the Social Security number. Efforts to match voter registrations in states that use different identifiers are complicated and may fail. Take, for example, the problem of figuring out whether Paul Smith in Michigan is the same person as Paul Smith in Kentucky. Since the unique identifier for voter registration is the driver’s license number in Michigan but the Social Security number in Kentucky, an accurate match of the two registered Paul Smiths is not likely. Any match will need to rely on Paul Smith’s date of birth to estimate, based on some level of probability, whether the Paul Smith in each state is the same person or not.
To make different state voter databases interoperable, therefore, they must use the same unique identifier, and this identifier must distinguish each American from every other voter in the country. The state voter databases will need to use a nationwide identifier. Since the same driver’s license number might be used in different states, the Social Security number provides the most feasible option for a federal unique identifier.
While the use of Social Security numbers for voter registration raises concerns about privacy, these concerns can be adequately addressed by the measures the Commission recommends to ensure the security of voter databases. The Commission stresses the importance for states to allow only authorized election officials to use the Social Security numbers. States should not provide Social Security numbers in the voter lists they release to candidates, political parties, or anyone else. This should not be hard to do. Forty-nine states collect Social Security numbers for driver’s licenses,12 and they have protected the privacy of the Social Security numbers. Congress should direct that all states use the same unique identifier — i.e., the voter’s Social Security number — and template, but a new system will also be needed to share data on voters among states. Such a system should maintain a uniform state voter list while allowing systematic updating of lists to take into account moves between states. The Commission proposes using a model similar to the one supervised by the U.S. Department of Transportation (DOT) to make sure that commercial drivers have only one license. The Commercial Driver’s License Information System (CDLIS) shares data among states on commercial driver’s licenses, using a “distributed database” — a collection of 51 databases (the 50 states and Washington, D.C.) that are linked to each other. When state officials want to check a particular driver’s record, they go to the central site, which then connects them to the database of the state that issued a commercial license to that particular driver. Since all of the state databases are inter-connected, an update in one state database is immediately available to all other states. CDLIS is operated by the American Association of Motor Vehicle Administrators under the supervision of the U.S. Department of Transportation.
Similarly, our Commission recommends a “distributed database” that will connect all states’ registration lists. The creation of a computerized system to transfer voter data between states is entirely feasible. This system could be managed either by the EAC or by an interstate compact or association of state officials under EAC supervision.
Implementation of the Commission’s recommendation on cross-state interoperability of voter databases will require state election authorities to collect Social Security numbers and digital images of signatures for all registered voters. While many states use the driver’s license number as their unique identifier, they can collect Social Security numbers from their state’s department of motor vehicles (a Social Security number is required by 49 states to issue a driver’s license).13
We recommend that the EAC oversee the adoption of the template for voter data and for assisting states in the creation of a new system to share voter data among states, including for setting up a distributed database.
Congress should appropriate federal funds to complete top-down state voter databases, cover the costs of adding Social Security numbers and digital images of signatures to the databases, and create and maintain the federal distributed database system for sharing voter data among states. Congress should provide these funds to the EAC for distribution to states that adopt the uniform template for voter data and join the system for data sharing. Federal funds would be withheld from states that do not make their voter files interoperable with the voter databases of other states.
As states make their voter databases interoperable, they will retain full control over their registration lists. They will only need to add to their current databases the voter data required to complete the uniform template.
Two additional innovations might help to eliminate registration problems that voters have encountered. First, voters should have an opportunity during the registration process and before Election Day to review the registration online list to see whether their name is correctly inscribed and to check their proper precinct for voting.14 Whenever an error is discovered, voters should notify the statewide registration office to correct it, and every statewide registration office should have procedures in place to correct such an error in a timely manner. Second, precincts should have an “electronic poll-book” that connects them to the statewide registration list and allows them to locate the correct polling site for each voter. For those precincts that are small, lack the resources for such an instrument, or do not have online access, precinct officials should telephone to a neighboring jurisdiction to obtain the correct information. Poll workers should also have a dedicated phone number to contact local election officials in case assistance is needed. This phone number should be different from the number provided to the public. Too often, poll workers cannot connect with election officials when assistance is needed because public phone lines are overwhelmed.
The entire system should permit state-of-the-art, computer-based registration lists that will be accurate and up-to-date for the entire nation.
2.2.1 In order to assure that lists take account of citizens moving from one state to another, voter databases should be made interoperable between states.This would serve to eliminate duplicate registrations, which are a source of potential fraud.
2.2.2 In order to assist the states in creating voter databases that are interoperable across states, the EAC should introduce a template for shared data and a format for crossstate data transfers.This template should include a person’s full legal name, date and place of birth, signature (captured as a digital image), and Social Security number.
2.2.3 With assistance and supervision by the EAC, a distributed database system should be established to make sure that the state lists remain current and accurate to take into account citizens moving between states. Congress should also pass a law mandating that states cooperate with this system to ensure that citizens do not vote in two states.
2.2.4 Congress should amend HAVA to mandate the interoperability of statewide registration lists. Federal funds should be appropriated for distribution by the EAC to states that make their voter databases interoperable, and the EAC should withhold federal funds from states that fail to do so.The law should also provide for enforcement of this requirement.
2.2.5 With proper safeguards for personal security, states should allow citizens to verify and correct the registration lists’ information on themselves up to 30 days before the election. States should also provide “electronic poll-books” to allow precinct officials to identify the correct polling site for voters.
2.2.6 With interoperability, citizens should need to register only once in their lifetime and updating their registration will be facilitated when they move.
Because of flaws in registration lists and other election administration procedures, HAVA mandated that any eligible voter who appears at the polls must be given a provisional ballot if his or her name does not appear on the voter registration list or an election official asserts that the individual is not eligible to vote. November 2, 2004, marked the first time that all states were supposed to offer provisional ballots in a general election. Out of 1.6 million provisional ballots cast, more than one million were counted.15 The 1.6 million provisional ballots do not include an unknown number of voters who were encouraged by poll workers to go to other polling sites where they might be registered.
Practices for offering and counting provisional ballots in the 2004 presidential election varied widely by state and by county. Around the country, the percentage of provisional ballots counted ranged from a national high in Alaska of 97 percent to a low of 6 percent in Delaware.16
This was due in part to whether a state accepted a provisional ballot cast outside of a voter’s home precinct. In other situations, provisional ballots were counted without first having been verified as eligible ballots.
If the recommendations for strengthening the registration lists are approved, the need for provisional ballots will be reduced. In 2004, provisional ballots were needed half as often in states with unified databases as in states without.17 Nonetheless, in the absence of the reforms recommended by this Commission, or in the period before they come fully into effect, provisional balloting will continue to be a crucial safety net. During the interim, in order to reduce the chances that elections are litigated, we need consistent procedures for handling provisional ballots and full training for poll workers who carry out these procedures.
2.3.1 Voters should be informed of their right to cast a provisional ballot if their name does not appear on the voter roll, or if an election official asserts that the individual is not eligible to vote, but States should take additional and effective steps to inform voters as to the location of their precinct.
2.3.2 States, not counties or municipalities, should establish uniform procedures for the verification and counting of provisional ballots, and that procedure should be applied uniformly throughout the State. Many members of the Commission recommend that a provisional ballot cast in the incorrect precinct but in the correct jurisdiction should be counted.
2.3.3 Poll workers should be fully trained on the use of provisional ballots, and provisional ballots should be distinctly marked and segregated so they are not counted until the eligibility of the voter is determined.
The hotlines set up by nonprofit organizations to assist voters on Election Day received hundreds of thousands of calls (see Table 1 on page 17). Most of the callers had two simple questions: Am I registered to vote? And where do I go to vote? Answers to these questions, however, too often were difficult to obtain. Only nine state election Web sites were able to provide voters with their registration information or with the address of their polling site. Information was equally difficult to obtain from election offices by telephone. One Election Day hotline transferred callers to their county board of elections, but barely half of these calls were answered, and of the other half, few provided the information that was requested.18
Failure to provide voters with such basic information as their registration status and their polling site location raises a barrier to voting as significant as inconsistent procedures on provisional ballots or voter ID requirements. As states gain responsibility for voter registration, they will be well positioned to inform voters if they are listed in the voter files. The Web sites of local jurisdictions should allow voters to check whether they are registered and the location of their precinct. This precinct-locator feature should be added to state elections Web sites. In addition, information on how to register and where to vote should be disseminated in local media, on posted lists, and in other government offices, including welfare and social services agencies.
Since election officials may have difficulty responding to telephone calls on Election Day as they are conducting the election, states and local jurisdictions should encourage voters to inquire about their registration status and the location of their polling place considerably before Election Day.
TABLE 1 : Voter Calls to the MYVOTE1 Hotline on Election Day 2004
|Topic of Question or Complaint on Election Day 2004||Percent of Total|
|Registration Issues/Poll Access||43.90%|
|NOTES: Totals are based upon an analysis of 55,000 phone calls to the MYVOTE1 hotline on November 2, 2004. Two major, nonpartisan hotlines and the U.S. Election Assistance Commission received a total of approximately 255,000 voter calls on Election Day 2004.|
|SOURCES: Testimony before the Commission on Federal Election Reform by Ken Smukler, President of Info Voter Technologies, on June 30, 2005; Testimony before the U.S. House of Representatives Administration Committee by the U.S. Election Assistance Commission, on February 9, 2005.|
2.4.1 States and local jurisdictions should use Web sites, toll-free numbers, and other means to answer questions from citizens as to whether they are registered and, if so, what is the location of their precinct, and if they are not registered, how they can do so before the deadline.
A good registration list will ensure that citizens are only registered in one place, but election officials still need to make sure that the person arriving at a polling site is the same one that is named on the registration list. In the old days and in small towns where everyone knows each other, voters did not need to identify themselves. But in the United States, where 40 million people move each year, and in urban areas where some people do not even know the people living in their own apartment building let alone their precinct, some form of identification is needed.
There is no evidence of extensive fraud in U.S. elections or of multiple voting, but both occur, and it could affect the outcome of a close election.19 The electoral system cannot inspire public confidence if no safeguards exist to deter or detect fraud or to confirm the identity of voters. Photo IDs currently are needed to board a plane, enter federal buildings, and cash a check. Voting is equally important.
The voter identification requirements introduced by HAVA are modest. HAVA requires only first-time voters who register by mail to show an ID, and they can choose from a number of different types of identification. States are encouraged to allow an expansive list of acceptable IDs, including those without a photograph, such as utility bills or government checks. These requirements were not implemented in a uniform manner and, in some cases, not at all. After HAVA was enacted, efforts grew in the states to strengthen voter identification requirements. While 11 states required voter ID in 2001, 24 states now require voters to present an ID at the polls.20 In addition, bills to introduce or strengthen voter ID requirements are under consideration in 12 other states.21
Our Commission is concerned that the different approaches to identification cards might prove to be a serious impediment to voting. There are two broad alternatives to this decentralized and unequal approach to identification cards. First, we could recommend eliminating any requirements for an ID because the evidence of multiple voting is thin, and ID requirements, as some have argued, are “a solution in search of a problem.” Alternatively, we could recommend a single national voting identification card. We considered but rejected both alternatives.
We rejected the first option — eliminating any requirements — because we believe that citizens should identify themselves as the correct person on the registration list when they vote. While the Commission is divided on the magnitude of voter fraud — with some believing the problem is widespread and others believing that it is minor — there is no doubt that it occurs. The problem, however, is not the magnitude of the fraud. In close or disputed elections, and there are many, a small amount of fraud could make the margin of difference. And second, the perception of possible fraud contributes to low confidence in the system. A good ID system could deter, detect, or eliminate several potential avenues of fraud— such as multiple voting or voting by individuals using the identities of others or those who are deceased — and thus it can enhance confidence. We view the other concerns about IDs — that they could disenfranchise eligible voters, have an adverse effect on minorities, or be used to monitor behavior — as serious and legitimate, and our proposal below aims to address each concern.
We rejected the second option of a national voting identification card because of the expense and our judgment that if these cards were only used for each election, voters would forget or lose them.
We therefore propose an alternative path. Instead of creating a new card, the Commission recommends that states use “REAL ID” cards for voting purposes. The REAL ID Act, signed into law in May 2005, requires states to verify each individual’s full legal name, date of birth, address, Social Security number, and U.S. citizenship before the individual is issued a driver’s license or personal ID card. The REAL ID is a logical vehicle because the National Voter Registration Act established a connection between obtaining a driver’s license and registering to vote. The REAL ID card adds two critical elements for voting — proof of citizenship and verification by using the full Social Security number.
The REAL ID Act does not require that the card indicates citizenship, but that would need to be done if the card is to be used for voting purposes. In addition, state bureaus of motor vehicles should automatically send the information to the state’s bureau of elections. (With the National Voter Registration Act, state bureaus of motor vehicles ask drivers if they want to register to vote and send the information only if the answer is affirmative.)
Reliance on REAL ID, however, is not enough. Voters who do not drive,22 including older citizens, should have the opportunity to register to vote and receive a voter ID. Where they will need identification for voting, IDs should be easily available and issued free of charge. States would make their own decision whether to use REAL ID for voting purposes or instead to rely on a template form of voter ID. Each state would also decide whether to require voters to present an ID at the polls, but our Commission recommends that states use the REAL ID and/or an EAC template for voting, which would be a REAL ID card without reference to a driver’s license.
For the next two federal elections, until January 1, 2010, in states that require voters to present ID at the polls, voters who fail to do so should nonetheless be allowed to cast a provisional ballot, and their ballot would count if their signature is verified. After the REAL ID is phased in, i.e., after January 1, 2010, voters without a valid photo ID, meaning a REAL ID or an EAC-template ID, could cast a provisional ballot, but they would have to return personally to the appropriate election office within 48 hours with a valid photo ID for their vote to be counted.
To verify the identity of voters who cast absentee ballots, the voter’s signature on the absentee ballot can be matched with a digitized version of the signature that the election administrator maintains. While such signature matches are usually done, they should be done consistently in all cases, so that election officials can verify the identity of every new registrant who casts an absentee ballot.
The introduction of voter ID requirements has raised concerns that they may present a barrier to voting, particularly by traditionally marginalized groups, such as the poor and minorities, some of whom lack a government-issued photo ID. They may also create obstacles for highly mobile groups of citizens. Part of these concerns are addressed by assuring that government-issued photo identification is available without expense to any citizen and, second, by government efforts to ensure that all voters are provided convenient opportunities to obtain a REAL ID or EAC-template ID card. As explained in Section 4.1, the Commission recommends that states play an affirmative role in reaching out with mobile offices to individuals who do not have a driver’s license or other government-issued photo ID to help them register to vote and obtain an ID card.
There are also longstanding concerns voiced by some Americans that national identification cards might be a step toward a police state. On that note, it is worth recalling that most advanced democracies have fraud-proof voting or national ID cards, and their democracies remain strong. Still, these concerns about the privacy and security of the card require additional steps to protect against potential abuse. We propose two approaches. First, new institutional and procedural safeguards should be established to assure people that their privacy, security, and identity will not be compromised by ID cards. The cards should not become instruments for monitoring behavior. Second, certain groups may see the ID cards as an obstacle to voting, so the government needs to take additional measures to register voters and provide ID cards.
The needed measures would consist of legal protections, strict procedures for managing voter data, and creation of ombudsman institutions. The legal protections would prohibit any commercial use of voter data and impose penalties for abuse. The data-management procedures would include background checks on all officials with access to voter data and requirements to notify individuals who are removed from the voter registration list. The establishment of ombudsman institutions at the state level would assist individuals to redress any cases of abuse. The ombudsman would be charged with assisting voters to overcome bureaucratic mistakes and hurdles and respond to citizen complaints about the misuse of data.
The Commission’s recommended approach to voter ID may need to adapt to changes in national policy in the future. Since the attacks of September 11, 2001, concerns about homeland security have led to new policies on personal identification. Under a presidential directive, about 40 million Americans who work for or contract with the federal government are being issued ID cards with biometrics, and the REAL ID card may very well become the principal identification card in the country. Driven by security concerns, our country may already be headed toward a national identity card. In the event that a national identity card is introduced, our Commission recommends that it be used for voting purposes as well.
2.5.1 To ensure that persons presenting themselves at the polling place are the ones on the registration list, the Commission recommends that states require voters to use the REAL ID card, which was mandated in a law signed by the President in May 2005. The card includes a person’s full legal name, date of birth, a signature (captured as a digital image), a photograph, and the person’s Social Security number.This card should be modestly adapted for voting purposes to indicate on the front or back whether the individual is a U.S. citizen. States should provide an EAC-template ID with a photo to non-drivers free of charge.
2.5.2 The right to vote is a vital component of U.S. citizenship, and all states should use their best efforts to obtain proof of citizenship before registering voters.
2.5.3 We recommend that until January 1, 2010, states allow voters without a valid photo ID card (Real or EAC-template ID) to vote, using a provisional ballot by signing an affidavit under penalty of perjury.The signature would then be matched with the digital image of the voter’s signature on file in the voter registration database, and if the match is positive, the provisional ballot should be counted. Such a signature match would in effect be the same procedure used to verify the identity of voters who cast absentee ballots. After January 1, 2010, voters who do not have their valid photo ID could vote, but their ballot would only count if they returned to the appropriate election office within 48 hours with a valid photo ID.
2.5.4 To address concerns about the abuse of ID cards, or the fear that it could be an obstacle to voting, states should establish legal protections to prohibit any commercial use of voter data and ombudsman institutions to respond expeditiously to any citizen complaints about the misuse of data or about mistaken purges of registration lists based on interstate matching or statewide updating.
2.5.5 In the event that Congress mandates a national identification card, it should include information related to voting and be connected to voter registration.
Voter registration lists provide the basis for determining who is qualified to vote. Yet only a few states, notably Oregon and North Carolina, have assessed the quality of their lists, or have developed plans to do so. This is also true as states rush to complete statewide voter databases before the January 1, 2006, deadline. Moreover, the EAC does not assess the quality of voter files.
The little information available on the quality of voter files is not reassuring. The creation of statewide voter databases allows for the elimination of duplicate registrations within states, but attempts to match voter files with records of other state agencies are often ineffective. Death records, for example, sometimes are not provided to election officials for three or four months, and information on felons is usually incomplete.23 Comparison with U.S. Census Bureau statistics also points to extensive “deadwood” on the voter registration lists. Some states have a large portion of inactive voters on their voter registration lists. One in four registered voters in Oregon is inactive, as is one in every three registered voters in California.24 There also are numerous jurisdictions, such as Alaska, where the number of registered voters is greater than the number of voting-aged citizens.25 These jurisdictions clearly have not updated their voter registration lists by removing the names of voters who have died or have moved away.
Voter registration lists are often inflated by the inclusion of citizens who have moved out of state but remain on the lists. Moreover, under the National Voter Registration Act, names are often added to the list, but counties and municipalities often do not delete the names of those who moved. Inflated voter lists are also caused by phony registrations and efforts to register individuals who are ineligible. Registration forms in the names of comic figures, for example, were submitted in Ohio in 2004. At the same time, inaccurate purges of voter lists have removed citizens who are eligible and are properly registered.
From what little is known, the quality of voter registration lists probably varies widely by state. Without quality assurance, however, cross-state transfers of voter data may suffer from the problem of “garbage in, garbage out.” They may pass on inaccurate data from certain states to the rest of the country. The overall quality of a system to share voter data among states will only be a strong as the quality of the worst state voter database.
Each state needs to audit its voter registration files to determine the extent to which they are accurate (with correct and current information on individuals), complete (including all eligible voters), valid (excluding ineligible voters), and secure (with protections against unauthorized use). This can be done by matching voter files with records in other state agency databases in a regular and timely manner, contacting individuals when the matches are inconclusive, and conducting survey research to estimate the number of voters who believe they are registered but who are not in fact listed in the voter files. Other countries regularly conduct such audits.26
Effective audits assess not only the quality of voter files but also the procedures used to update, maintain, and verify data and to ensure security of voter databases. To assure continual quality of voter databases, effective procedures are needed to maintain up-to-date lists of eligible voters, verify the accuracy of those lists, and remove voters who have become ineligible. These should include procedures to delete those who have moved out of state and to effectively match voter files with records of driver’s licenses, deaths, and felons. Given the controversial “purges” that have occurred, special care must be taken to update the lists in a fair and transparent manner. States should adopt uniform procedures and strong safeguards against incorrect removal of eligible voters. Every removal should be doublechecked before it is executed, and a record should be kept of every action. The process of updating the lists should be continuous, and before each statewide election the voter rolls should be audited for accuracy.
In addition, states need to assure the privacy and security of voter files. There is no justification for states to release voter files for commercial purposes. However, components of voter files should remain public documents subject to public scrutiny. States must carefully balance the right to privacy of registered citizens with the need for transparency in elections when they decide what information on voter registration to make available to the public. Procedures are also needed to protect voter files against tampering or abuse. This might be done by setting up the voter database to make an automatic record of all changes to the voter files, including a record of who made the changes and when.
2.6.1 States need to effectively maintain and update their voter registration lists.The EAC should provide voluntary guidelines to the states for quality audits to test voter registration databases for accuracy (correct and up-to-date information on individuals), completeness (inclusion of all eligible voters), and security (protection against unauthorized access). When an eligible voter moves from one state to another, the state to which the voter is moving should be required to notify the state which the voter is leaving to eliminate that voter from its registration list.
2.6.2 All states should have procedures for maintaining accurate lists such as electronic matching of death records, drivers licenses, local tax rolls, and felon records.
2.6.3 Federal and state courts should provide state election offices with the lists of individuals who declare they are non-citizens when they are summoned for jury duty.
2.6.4 In a manner that is consistent with the National Voter Registration Act, states should make their best efforts to remove inactive voters from the voter registration lists. States should follow uniform and strict procedures for removal of names from voter registration lists and should adopt strong safeguards against incorrect removal of eligible voters. All removals of names from voter registration lists should be double-checked.
2.6.5 Local jurisdictions should track and document all changes to their computer databases, including the names of those who make the changes.
^ Back to top
The Help America Vote Act of 2002 authorized up to $650 million in federal funds to replace antiquated voting machines throughout the country. States are using these funds and their own resources to upgrade voting technology, generally to replace punch card and lever voting machines with new optical scan and electronic voting systems. As a result, voting technology is improving,27 but new concerns related to electronic voting systems have arisen. These concerns need to be addressed, because it is vital to the electoral process that citizens have confidence that voting technologies are registering and tabulating votes accurately.
The purpose of voting technology is to record and tally all votes accurately and to provide sufficient evidence to assure all participants — especially the losing candidates and their supporters — that the election result accurately reflects the will of the voters.
Voting machines must be both accessible and transparent. As required by HAVA, the machines must be accessible to language minorities and citizens with disabilities, including the blind and visually impaired citizens, in a manner that allows for privacy and independence. Voting machines must also be transparent. They must allow for recounts and for audits, and thereby give voters confidence in the accuracy of the vote tallies.
Two current technology systems are optical scan and direct recording electronic (DRE) systems. Optical scan systems rely on preprinted paper ballots that are marked by the voter, like the ovals students fill in with a No. 2 pencil on a standardized exam, and then are run through an optical scan machine that determines and tallies the votes. Such systems provide transparency because the paper ballots can be recounted and audited by hand. Under HAVA, all aspects of the voting system, including the production of audit trail information, must be accessible to voters with disabilities.
DRE machines present voters with their choices on a computer screen, and voters choose by touching the screen or turning a dial. The vote is then recorded electronically, usually without ballot paper. DREs make up a growing share of voting equipment. Nearly 30 percent of voters live in jurisdictions that use DREs, compared to 17 percent in the 2000 election (see Table 2 on page 27).28 DREs allow voters with disabilities to use audio prompts to cast ballots privately and independently, and they facilitate voting by non-English speakers by offering displays of the ballot in different languages. DREs also provide greater accuracy in recording votes, in part by preventing over-votes, whereby people mistakenly vote for more than one candidate, and by discouraging accidental under-votes by reminding voters when they overlooked one or more races.
The accessibility and accuracy of DREs, however, are offset by a lack of transparency, which has raised concerns about security and verifiability. In most of the DREs used in 2004, voters could not check that their ballot was recorded correctly. Some DREs had no capacity for an independent recount. And, of course, DREs are computers, and computers malfunction. A malfunction of DREs in Carteret County, North Carolina, in the November 2004 elections caused the loss of more than 4,400 votes. There was no backup record of the votes that were cast. As a result, Carteret County had no choice but to rerun the election, after which it abandoned its DREs. Other jurisdictions have lost votes because election officials did not properly set up voting machines.29
To provide backup records of votes cast on DREs, HAVA requires that all voting machines produce a “permanent paper record with a manual audit capacity.” This requirement is generally interpreted to mean that each machine must record individual ballot images, so that they can be printed out and examined in the event of a disputed result or of a recount. This will make DREs somewhat more transparent, but it is still insufficient to fully restore confidence.
One way to instill greater confidence that DREs are properly recording votes is to require a paper record of the ballot that the voter can verify before the ballot is cast. Such a paper record, known as a voter-verifiable paper audit trail (VVPAT), allows the voter to check that his or her vote was recorded as it was intended.
Because voter-verifiable paper audit trails can permit recounts, audits, and a backup in case of a malfunction, there is a growing demand for such paper trails. As of early August 2005, 25 states required voter-verifiable paper ballots, and another 14 states had proposed legislation with such a requirement.30
Since very few of the DREs in use today are equipped to print voter-verifiable paper audit trails, certain bills before Congress would require election authorities to “retrofit” DREs with such printers. In 2004, DREs with voter-verifiable paper audit trails were used only in Nevada. They appear to have worked well.31 When Nevadans went to the polls and made their selection, a paper record of their vote was printed behind a glass cover on a paper roll, like the roll of paper in a cash register. Voters were able to view the paper record and thereby check that their vote was recorded accurately before they cast their ballot. The paper record was saved in the machine and thus was available for later use in recounts or audits. After the 2004 elections, Nevada election officials conducted an internal audit, which confirmed the accuracy of the votes recorded by the DREs. While less than one in three Nevada voters reportedly looked at the paper record of their ballot, these voters had the opportunity to confirm their vote, and the paper allowed a chance to verify the computer tallies after the election.
While HAVA already requires that all precincts be equipped with at least one piece of voting equipment that is fully accessible to voters with disabilities for use in federal elections by January 1, 2006, must be accessible to voters with disabilities, the Commission believes that transparency in voting machines should also be assured in time for the 2008 presidential election. With regard to current technology, states will need to use either DREs with a voter-verifiable paper audit trail and an audio prompt for blind voters or optical scan voting systems with at least one computer-assisted marking device for voters with disabilities to mark their ballot. To ensure implementation of this requirement, Congress will need to appropriate sufficient funds to cover the costs of either retrofitting DREs with voterverifiable paper audit trails or purchasing a computer-assisted marking device for each polling place that uses optical scan voting systems.
Concerns have been raised that the printers could malfunction just as computers do. Of course, the previous ballot papers will be available, and the operators will know when the printers fail. Still, precincts should have backup printers for that contingency. A second concern is that the length of the ballot in some areas — such as California, which frequently has referenda — would require paper trails that would be several feet long. In the case of non-federal races, state law would determine whether the non-federal portion of the ballot would similarly be required to provide a voter-verified paper audit trail. That is not a perfect solution, but it is still better than having no paper backup at all.
The standards for voting systems, set by the EAC, should assure both accessibility and transparency in all voting machines. Because these standards usually guide the decisions of voting machine manufacturers, the manufacturers should be encouraged to build machines in the future that are both accessible and transparent and are fully capable of meeting the needs of Americans with disabilities, of allowing voters to verify their ballots, and of providing for independent audits of election results.
|TABLE 2: Types of Voting Equipment Used in Recent Presidential Elections|
|Type of Voting||Registered Voters in 2000||Registered Voters in 2004|
|SOURCE: Election Data Services, Voting Equipment Summary by Type, 2004. Election Data Services, New Study Shows 50 Million Voters Will Use Electronic Voting Systems, 32 with Punch Cards in 2004.|
3.1.1 Congress should pass a law requiring that all voting machines be equipped with a voter-verifiable paper audit trail and, consistent with HAVA, be fully accessible to voters with disabilities.This is especially important for direct recording electronic (DRE) machines for four reasons: (a) to increase citizens’ confidence that their vote will be counted accurately, (b) to allow for a recount, (c) to provide a backup in cases of loss of votes due to computer malfunction, and (d) to test — through a random selection of machines — whether the paper result is the same as the electronic result. Federal funds should be appropriated to the EAC to transfer to the states to implement this law.While paper trails and ballots currently provide the only means to meet the Commission’s recommended standards for transparency, new technologies may do so more effectively in the future.The Commission therefore urges research and development of new technologies to enhance transparency, security, and auditability of voting systems.
3.1.2 States should adopt unambiguous procedures to reconcile any disparity between the electronic ballot tally and the paper ballot tally.The Commission strongly recommends that states determine well in advance of elections which will be the ballot of record.
While voter-verifiable paper ballots will contribute to strengthening public confidence in DREs, regular audits of voting machines are also needed to double-check the accuracy of the machines’ vote tallies. Such audits were required by law in 10 states as of mid-August 2005.32 To carry out such audits, election officials would randomly select a sample of voting machines and compare the vote total recorded by the machines with the vote total on the paper ballots. The audits would test the reliability of voting machines and identify problems, often before a close or disputed election takes place. This, in turn, would encourage both suppliers and election officials to effectively maintain voting machines.
Some concern has been expressed about the possibility of manipulation of paper audit trails.33 If DREs can be manipulated to alter the vote tallies, the same can be done with paper audit trails. Such manipulation can be detected and deterred by regular audits of voting machines. Regular audits should be done of all voting machines, including DREs and optical scan systems.
3.2.1 State and local election authorities should publicly test all types of voting machines before, during, and after Election Day and allow public observation of zero machine counts at the start of Election Day and the machine certification process.
DREs run on software that can be compromised. DRE software may get attacked or hacked by outsiders, perhaps through the Internet. As experience in computer security shows, it is often difficult to defend against such attacks. Hackers often are creative and determined, and voting systems provide a tempting target. However, while some DREs send their results to election headquarters over the Internet, they are not connected to the Internet during voting.
The greater threat to most systems comes not from external hackers, but from insiders who have direct access to the machines. Software can be modified maliciously before being installed into individual voting machines. There is no reason to trust insiders in the election industry any more than in other industries, such as gambling, where sophisticated insider fraud has occurred despite extraordinary measures to prevent it. Software can also be programmed incorrectly. This poses a likely threat when local programmers who lack the necessary skills nonetheless modify the ballot for local offices, and many might not have the sophistication required for the new machines.
In addition to the output of DREs, which can be verified through a paper audit trail, the inside process of programming DREs should be open to scrutiny by candidates, their supporters, independent experts, and other interested citizens, so that problems can be detected, deterred, or corrected, and so that the public will have confidence in the machines.
At the same time, manufacturers of voting machines have legitimate reason to keep their voting machine software and its source code proprietary. The public interest in transparency and the proprietary interests of manufacturers can be reconciled by placing the source code in escrow with the National Institute of Standards and Technology (NIST), and by making the source code available for inspection on a restricted basis to qualified individuals. NIST might make the source code available to recognized computer security experts at accredited universities and to experts acting on behalf of candidates or political parties under a nondisclosure agreement, which would bar them from making information about the source code public, though they could disclose security flaws or vulnerabilities in the voting system software.
Doubt has been raised that some manufacturers of voting machines provide enough security in their systems to reduce the risk of being hacked. Such concerns were highlighted after a group of computer security experts examined a voting system source code that was accidentally left on the Internet.34 Independent inspection of source codes would strengthen the security of voting systems software by encouraging manufacturers to improve voting system security. Expert reviews may also detect software design flaws or vulnerabilities. This, in turn, could bolster public confidence in the reliability of DREs to accurately record and tally the vote in elections.
In addition to the source codes, the software and the voting machines themselves are potentially vulnerable to manipulation. Security for voting systems should guard against attempts to tamper with software or individual voting machines. When voting machines are tested for certification, a digital fingerprint, also known as a “hash,” of their software is often sent to NIST. Following the delivery of new voting machines, a local jurisdiction can compare the software on these machines to the digital fingerprint at NIST. This comparison either will identify changes made to the software before delivery or, if the software is unaltered, will confirm that the software on the individual machines meets the certified standards.
Once voting machines arrive at the local jurisdiction, election officials must take precautions to ensure security by restricting access to authorized personnel and by documenting access to the machines.
The process of testing and certifying voting machines is designed mainly to ensure their reliability. Testing and certification is conducted under EAC supervision, although some states require additional testing and certification. The state testing can make the process more rigorous, particularly when voting machines are field tested. When California conducted a mock election with new voting machines in July 2005, it found unacceptable rates of malfunctions that were not apparent in lab tests.35
No matter how secure voting machines are or how carefully they are used, they are liable to malfunction. To avoid a situation where a machine malfunction will cause a major disruption, local jurisdictions need to prepare for Election Day with a backup plan, including how the vendor will respond to a machine malfunction and what alternatives, including paper ballots, should be made available.
3.3.1 The Independent Testing Authorities, under EAC supervision, should have responsibility for certifying the security of the source codes to protect against accidental or deliberate manipulation of vote results. In addition, a copy of the source codes should be put in escrow for future review by qualified experts. Manufacturers who are unwilling to submit their source codes for EAC-supervised testing and for review by independent experts should be prohibited from selling their voting machines.
3.3.2 States and local jurisdictions should verify upon delivery of a voting machine that the system matches the system that was certified.
3.3.3 Local jurisdictions should restrict access to voting equipment and document all access, as well as all changes to computer hardware or software.
3.3.4 Local jurisdictions should have backup plans in case of equipment failure on Election Day.
The Internet has become such a pervasive influence on modern life that it is natural for the public and election officials to begin considering ways to use it to facilitate voting. The first binding Internet election for political office took place in 2000, when the Arizona Democratic Party used it during its primary. In 2004, the Michigan Democratic Party allowed voting by Internet during its caucuses. Meanwhile, Missouri announced that any member of the U.S. military serving in combat areas overseas could complete an absentee ballot for the general election and email a scanned copy to the Department of Defense, which then would forward it to the appropriate local election offices.
Despite these much-publicized trials, serious concerns have been raised about the push for a “digital democracy.” In 2004, the Department of Defense cancelled its $22 million Secure Electronic and Voting Registration Experiment (SERVE) program designed to offer Internet voting during the presidential election to members of the U.S. military and other overseas citizens. The cancellation came after a group of top computer scientists who reviewed the system reported that without improved security, Internet voting is highly susceptible to fraud.
First, there are the issues of privacy and authentication. When using the Internet, one cannot assure voters that their ballot will remain secret. Second, the current system is not fully secure. Although data sent via the Internet can be encrypted and then decoded by local election administrators, hackers can compromise the system. This was the conclusion of the computer scientists who reviewed the SERVE program for the Pentagon. Due to security threats, some state and local election offices do not allow vote totals to be transmitted via the Internet. Third, no government or industry standards specifically apply to Internet voting technology. The EAC may begin developing such standards, but that work has not begun. Finally, Internet voting from homes and offices may not provide the same level of privacy as the voting booth.
To date, the most comprehensive study of Internet voting is contained in a 2001 report sponsored by the National Science Foundation.36 This report urges further research and experimentation to deal with the problems posed by this form of voting. Its authors suggest that it will take at least a decade to examine the various security and authentication issues. Our Commission agrees that such experimentation is necessary, and that the time for Internet voting has not yet arrived.
^ Back to top
The Commission believes that the vitality of America’s democracy depends on the active participation of our citizens. Yet, even in the presidential election in 2004, when voter interest was higher than normal, more than one in three eligible voters did not participate. We need to do more to increase voter participation, and we have considered numerous methods. None of them will solve the problem, but we encourage states to experiment with alternatives to raise the level of voter participation.
Recent elections have seen a substantial increase in early voting and in voting by mail. While only 8 percent of ballots were cast before Election Day in 1994, by 2004 the percentage of ballots cast before Election Day had risen to 22 percent. This increase in early and convenience voting has had little impact on voter turnout, because citizens who vote early or vote by mail tend to vote anyway.37 Early and convenience voting are popular, but there is little evidence that they will significantly expand participation in elections.38
There are other measures that can be taken to expand participation, particularly for military and overseas voters and for citizens with disabilities. There is also much to do with regard to civic and voter education that could have a long-term and lasting effect, particularly on young people. However, we first need to reach out to all eligible voters and remove any impediments to their participation created by the registration process or by identification requirements.
All citizens, including citizens with disabilities, need to have access to polling places. Polling places should be located in public buildings and other semipublic venues such as churches and community centers that comply with the Americans with Disability Act (ADA). Additionally, polling places should be located and protected so that voters can participate free of intimidation and harassment. Polling places should not be located in a candidate’s headquarters or in homes or business establishments that are not appropriately accessible to voters with disabilities.
The Commission’s proposals for a new electoral system contain elements to assure the quality of the list and the integrity of the ballot. But to move beyond the debate between integrity and access, specific and important steps need to be taken to assure and improve access to voting.
States have a responsibility to make voter registration accessible by taking the initiative to reach out to citizens who are not registered, for instance by implementing provisions of the National Voter Registration Act that allow voter registration at social-service agencies or by conducting voter registration and REAL ID card drives with mobile offices. Michigan, for example, uses a mobile office to provide a range of services, including driver’s licenses and voter registration. This model should be extended to all the states.
Political party and nonpartisan voter registration drives generally contribute to the electoral process by generating interest in upcoming elections and expanding participation. However, they are occasionally abused. There were reports in 2004 that some party activists failed to deliver voter registration forms of citizens who expressed a preference for the opposing party. During the U.S. House Administration Committee hearings in Ohio, election officials reported being deluged with voter registration forms at the last minute before the registration deadline, making it difficult to process these registrations in a timely manner. Many of the registration forms delivered in October to election officials were actually collected in the spring.
Each state should therefore oversee political party and nonpartisan voter registration drives to ensure that they operate effectively, that registration forms are delivered promptly to election officials, that all completed registration forms are delivered to the election officials, and that none are “culled” and omitted according to the registrant’s partisan affiliation. Measures should also be adopted to track and hold accountable those who are engaged in submitting fraudulent voter registrations. Such oversight might consist of training activists who conduct voter registration drives and tracking voter registration forms to make sure they are all accounted for. The tracking of voter registration forms will require better cooperation between the federal and state governments, perhaps through the EAC, as the federal government puts some registration forms online. In addition, states should apply a criminal penalty to any activist who deliberately fails to deliver a completed voter registration form.
4.1.1 States should undertake their best efforts to make voter registration and ID accessible and available to all eligible citizens, including Americans with disabilities. States should also remove all unfair impediments to voter registration by citizens who are eligible to vote.
4.1.2 States should improve procedures for voter registration efforts that are not conducted by election officials, such as requiring state or local registration and training of any “voter registration drives.”
4.1.3 Because there have been reports that some people allegedly did not deliver registration forms of those who expressed a preference for another party, states need to take special precautions to assure that all voter registration forms are fully accounted for. A unique number should be printed on the registration form and also on a detachable receipt so that the voter and the state election office can track the status of the form.39 In addition, voter registration forms should be returned within 14 days after they are signed.
A growing number of Americans vote by mail. Oregon moved entirely to a vote-by-mail system in 1998, and the practice of casting ballots by mail has continued to expand nationwide as voters and election officials seek alternatives to the traditional system of voting at polling stations. The state legislatures of California and of Washington state have considered legislation to expand the use of vote by mail, and in 24 states no excuse is required to vote absentee.
The impact of vote by mail is mixed. Proponents argue that vote by mail facilitates participation among groups that experience low voter turnout, such as elderly Americans and Native Americans.
While vote by mail appears to increase turnout for local elections, there is no evidence that it significantly expands participation in federal elections.40 Moreover, it raises concerns about privacy, as citizens voting at home may come under pressure to vote for certain candidates, and it increases the risk of fraud. Oregon appears to have avoided significant fraud in its vote-by-mail elections by introducing safeguards to protect ballot integrity, including signature verification. Vote by mail is, however, likely to increase the risks of fraud and of contested elections in other states, where the population is more mobile, where there is some history of troubled elections, or where the safeguards for ballot integrity are weaker.
The case of King County, Washington, is instructive. In the 2004 gubernatorial elections, when two in three ballots there were cast by mail, authorities lacked an effective system to track the number of ballots sent or returned. As a result, King County election officials were unable to account for all absentee ballots. Moreover, a number of provisional ballots were accepted without signature verification.41 The failures to account for all absentee ballots and to verify signatures on provisional ballots became issues in the protracted litigation that followed Washington state’s 2004 gubernatorial election.
Vote by mail is popular but not a panacea for declining participation. While there is little evidence of fraud in Oregon, where the entire state votes by mail, absentee balloting in other states has been one of the major sources of fraud. Even in Oregon, better precautions are needed to ensure that the return of ballots is not intercepted.
The evidence on “early” voting is similar to that of vote by mail. People like it, but it does not appear to increase voter participation, and there are some drawbacks. It allows a significant portion of voters to cast their ballot before they have all of the information that will become available to the rest of the electorate. Crucial information about candidates may emerge in the final weeks or even days of an election campaign. Early and convenience voting also detracts from the collective expression of citizenship that takes place on Election Day. Moreover, the cost of administering elections and of running campaigns tends to increase when early and mail-in voting is conducted in addition to balloting on Election Day. Early voting should commence no earlier than 15 days prior to the election, so that all voters will cast their ballots on the basis of largely comparable information about the candidates and the issues.
4.2.1 The Commission encourages further research on the pros and cons of vote by mail and of early voting.
Another alternative to voting at polling stations is the innovation of “vote centers,” pioneered by Larimer County, Colorado. Vote centers are larger in size than precincts but fewer in number. They are dispersed throughout the jurisdiction, but close to heavy traffic routes, larger residential areas, and major employers. These vote centers allow citizens to vote anywhere in the county rather than just at a designated precinct. Because these vote centers employ economies of scale, fewer poll workers are required, and they tend to be more professional. Also, the vote centers are reported to use more sophisticated technology that is more accessible to voters with disabilities. Vote centers eliminate the incidence of out-of-precinct provisional ballots, but they need to have a unified voter database that can communicate with all of the other centers in the county to ensure that eligible citizens vote only once.
While vote centers appear to have operated effectively in Larimer County, further research is needed to determine if the costs of establishing vote centers are offset by the savings of eliminating traditional polling sites. Moreover, because vote centers replace traditional voting at precincts, which are generally closer to a voter’s home, it is not clear that citizens actually view them as more convenient.
4.3.1 States should modify current election law to allow experimentation with voting centers. More research, however, is needed to assess whether voting centers expand voter participation and are cost effective.
4.3.2 Voting centers need a higher quality, computer-based registration list to assure that citizens can vote at any center without being able to vote more than once.
Military and overseas voting present substantial logistical challenges, yet we cannot overstate the imperative of facilitating participation in elections by military and overseas voters, particularly by service men and women who put their lives on the line for their country. The Commission calls on every state, with federal government assistance, to make every effort to provide all military and overseas voters with ample opportunity to vote in federal elections.
More than six million eligible voters serve in the Armed Forces or live overseas. These voters include 2.7 million military and their dependents and 3.4 million diplomats, Peace Corps volunteers, and other civilian government and other citizens overseas.42
Voter turnout among members of the armed forces is high. So is the level of frustration they experience when their votes cannot be counted. This happens largely because of the time required by the three-step process of applying for an absentee ballot, receiving one, and then returning a completed ballot. The process is complicated by the differences among states and among localities in the registration deadline, ballot format, and requirements for ballot return, and it is exacerbated because of the mobility of service men and women during a time of conflict. Since September 11, 2001, more than 500,000 National Guard and Reserve personnel have been mobilized, and many were relocated before they received their absentee ballots.
Congress passed the Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA) in 1986 to help eligible members of the armed services and their families, and other citizens overseas, to vote. UOCAVA required each state to have a single office to provide information on voter registration and absentee ballot procedures for military voters. The Help America Vote Act of 2002 (HAVA) recommended — but did not require — that this state office should coordinate voting by military personnel by receiving absentee ballot applications and collecting voted ballots. The introduction of statewide voter registration databases under HAVA provides an opportunity to put this recommendation into practice. But aside from Alaska, which already had a single state office, no state has centralized the processing of absentee ballots. This is another example as to why recommending, rather than requiring, a course of action is insufficient.
The Commission recommends that when registering members of the armed forces and other overseas voters, states should inquire whether to send an absentee ballot to them automatically, thus saving a step in the process.
In the 2004 presidential election, approximately one in four military voters did not vote for a variety of reasons: The absentee ballots were not returned or arrived too late; they were rejected for procedural deficiencies, such as a signature not properly witnessed on the back of the return envelope; blank ballots were returned as undeliverable; or Federal Post Card Applications were rejected.43
The U.S. Department of Defense’s Federal Voting Assistance Program, which assists military and overseas voters, tried to reduce the time lag for absentee voting by launching an electronic voting experiment. However, this experiment was ended because of fundamental security problems (see above on “Internet voting”).44 In the meantime, the Federal Voting Assistance Program encouraged states to send blank ballots out electronically and to accept voted ballots by fax. There now are 32 states that permit fax delivery of a blank ballot to military voters and 25 states that allow military voters to return their voted ballot by fax. In addition, some jurisdictions allow the delivery of blank ballots by email.45 The return of voted ballots by fax or email, however, is a violation of the key principle of a secret ballot, and it is vulnerable to abuse or fraud.
Although the Uniformed and Overseas Citizens Absentee Voting Act applies to both military and nonmilitary voters overseas, procedures to facilitate overseas voting serve military voters better than civilians. To provide civilian overseas voters with equal opportunities to participate in federal elections, new approaches are needed at both the federal and state levels.
4.4.1 The law calling for state offices to process absentee ballots for military and overseas government and civilian voters should be implemented fully, and these offices should be under the supervision of the state election offices.
4.4.2 New approaches should be adopted at the federal and state levels to facilitate voting by civilian voters overseas.
4.4.3 U.S. Department of Defense (DOD) should supply to all military posted outside the United States a Federal Postcard Application for voter registration and a Federal Write-in Absentee Ballot for calendar years in which there are federal elections.With adequate security protections, it would be preferable for the application forms for absentee ballots to be filed by Internet.
4.4.4 The states, in coordination with the U.S. Department of Defense’s Federal Voting Assistance Program, should develop a system to expedite the delivery of ballots to military and overseas civilian voters by fax, email, or overnight delivery service, but voted ballots should be returned by regular mail, and by overnight mail whenever possible.The Defense Department should give higher priority to using military aircraft returning from bases overseas to carry ballots.Voted ballots should not be returned by email or by fax as this violates the secrecy of the ballot and is vulnerable to fraud.
4.4.5 All ballots subject to the Uniform and Overseas Civilians Absentee Voting Act must be mailed out at least 45 days before the election (if request is received by then) or within two days of receipt after that. If the ballot is not yet set, due to litigation, a late vacancy, etc., a temporary ballot listing all settled offices and ballot issues must be mailed.
4.4.6 States should count the ballots of military and overseas voters up to 10 days after an election if the ballots are postmarked by Election Day.
4.4.7 As the technology advances and the costs decline, tracking systems should be added to absentee ballots so that military and overseas voters may verify the delivery of their voted absentee ballots.
4.4.8 The Federal Voting Assistance Program should receive a copy of the report that states are required under HAVA to provide the EAC on the number of absentee ballots sent to and received from military and overseas voters.
There are almost 30 million voting-aged Americans with some kind of disability—about 15 percent of the population (see Table 3 on page 40). Less than half of them vote. There are federal laws to facilitate voting and registration by eligible Americans with disabilities, but these laws have not been implemented with any vigor. As a result, voters with disabilities still face serious barriers to voting.46 Congress passed the Voting Accessibility for the Elderly and Handicapped Act in 1984 and the Americans with Disabilities Act of 1990, which required local authorities to make polling places physically accessible to people with disabilities for federal elections. Yet a Government Accountability Office survey of the nation’s polling places in 2000 found that 84 percent of polling places were not accessible on Election Day. By 2004, accessibility for voters with disabilities had improved only marginally. Missouri, for example, surveyed every polling place in the state and found that 71 percent were not accessible. Most other states have not even conducted surveys.47
There is similarly weak implementation of laws designed to facilitate voter registration by citizens with disabilities. Section 7 of the National Voter Registration Act (NVRA) requires state-funded agencies which provide services to citizens with disabilities to offer the opportunity to register citizens to vote. Implementation of this requirement, according to advocates for voters with disabilities, is rare or poor.48
HAVA provided additional support to Section 7 of NVRA by including social-service agencies as places to register voters, but only one state, Kentucky, has complied with Section 7, according to advocates for voters with disabilities. Moreover, at the current time, there is not a single case where the new statewide voter databases comply with Section 7.49 Thus, 12 years after the National Voter Registration Act was passed, voters with disabilities still cannot apply for voter registration at all social service offices.
|TABLE 3: Estimates of U.S.Voting Population with Disabilities by Type|
Age 16 and Older
|Percent of Total
|Sensory, Physical, Mental or Self-Care Disability||29.5||15%|
|Sensory and Physical Disability||2.5||1%|
|Sensory, Physical, and Mental Disability||2||1%|
|Total Voting Age Population in the U.S. (18 and older)||203||100%|
|NOTES: Respondents were able to report more than one type of disability.
SOURCES: U.S. Census Bureau, Selected Types of Disability for the Civilian Noninstitutionalized Population 5 Years and Over by Age:
2000; U.S. Census Bureau, Voting and Registration in the Election of November 2000.
4.5.1 To improve accessibility of polling places for voters with disabilities, the U.S. Department of Justice should improve its enforcement of the Americans with Disabilities Act and the accessibility requirements set by the Help America Vote Act.
4.5.2 States should make their voter registration databases interoperable with social-service agency databases and facilitate voter registration at social-service offices by citizens with disabilities.
4.5.3 States and local jurisdictions should allow voters with disabilities to request an absentee ballot when they register and to receive an absentee ballot automatically for every subsequent election. Local election officials should determine which voters with disabilities would qualify.
Only Maine and Vermont allow incarcerated citizens to vote. In all other states, citizens who are convicted of a felony lose their right to vote, either temporarily or permanently. An estimated 4.65 million Americans have currently or permanently lost their right to vote as a result of a felony conviction. Most states reinstate that right upon completion of the full sentence, including of parole, but three states — Florida, Kentucky, and Virginia — permanently ban all ex-felons from voting, and another 10 states have a permanent ban on voting by certain categories of ex-felons.50 These laws have a disproportionate impact on minorities.
Some states impose a waiting period after felons complete their sentence before they can vote. Few states take the initiative to inform ex-felons when their voting rights are restored. As a result, only a small portion of the ex-felons who have regained their voting rights are registered to vote.
Proponents of re-enfranchisement argue that ex-felons have paid their debt to society when they have completed their full sentence. Restoring their right to vote would encourage them to reintegrate into society. Each state therefore should automatically restore the voting rights of ex-felons who have completed their full sentence, including any terms of parole and compensation to victims. Opponents of re-enfranchisement, however, see this as a “punishment” issue rather than a “voting rights” issue. They believe that each state should be free to decide whether to restore the voting rights of ex-felons. States set punishment for state crimes, and this often extends beyond the completion of a felon’s sentence. Ex-felons are, for instance, usually barred from purchasing firearms or from getting a job as a publicschool teacher. Nonetheless, weighing both sides of the debate, the Commission believes that voting rights should be restored to certain categories of felons after they served the debt to society.
Recommendations on Re-Enfranchisement of Ex-Felons
4.6.1 States should allow for restoration of voting rights to otherwise eligible citizens who have been convicted of a felony (other than for a capital crime or one which requires enrollment with an offender registry for sex crimes) once they have fully served their sentence, including any term of probation or parole.
4.6.2 States should provide information on voter registration to ex-felons who have become eligible to vote. In addition, each state’s department of corrections should automatically notify the state election office when a felon has regained eligibility to vote.
Among the simplest ways to promote greater and more informed participation in elections is to provide citizens with basic information on voting and the choices that voters will face in the polling booth. HAVA requires only that basic voter information, including a sample ballot and instructions on how to vote, be posted at each polling site on Election Day. However, additional voter information is needed.
States or local jurisdictions should provide information by mail and on their Web sites to educate voters on the upcoming ballot — on the issues and the candidates, who will provide the information about themselves. Local election officials should set limits on the amount — but not the content — of information to be provided by the candidates. In Washington state, for example, every household is mailed a pamphlet with information on how to register, where to vote, and texts of election laws and proposed ballot initiatives and referendums. This voter’s pamphlet also has a picture of each candidate for statewide office and a statement of the candidate’s goals for the office they seek. In addition, there should be greater use of the radio and television to communicate these messages.
Efforts to provide voter information and education to young Americans merit particular attention. Voter turnout among youth declined steadily from the 1970s to 2000, when it was 24 percent lower than turnout of the entire electorate. In 2004, however, there was a surge of 11 percent in voter turnout among Americans aged 18 to 24, and the gap between youth turnout and overall turnout dropped to 17 percent (see Table 4).51
While participation by youth increased significantly in the last election, it continues to lag far behind the rest of the population. It can and should be increased by instructing high school students on their voting rights and civic responsibilities. Just one course in civics or American government can have a strong influence on youth participation in elections. According to a 2003 survey, about twice as many young Americans who have taken a civics course are registered to vote and have voted in all or most elections than young Americans who have never taken such a course.52
Moreover, Americans want public schools to prepare their children for citizenship and to provide better civic education. While most Americans believe that the most important goal of public schools is to develop basic skills, seven in 10 respondents to a 2004 survey agreed that preparing students to become responsible citizens is a “central purpose of public schools.” When asked to grade the civic education programs of public schools, 54 percent of respondents give these programs a “C” and 22 percent give them a “D.”53
It is difficult to assess the current efforts of state and local voting and civic education programs because only one state, Florida, publishes a report on its activities and spending in this area. We recommend that more states and local jurisdictions follow Florida’s example in order to generate more information on the most effective methods for voter and civic education.
TABLE 4: Voter Turnout in Presidential Elections by Age, 1972-2004
|18 to 24 years||49.6||42.2||39.9||40.8||36.2||42.8||32.4||32.3||41.9|
|25 to 44 years||62.7||58.7||58.7||58.4||54||58.3||49.2||49.8||52.2|
|45 to 64 years||70.8||68.7||69.3||69.8||67.9||70||64.4||64.1||66.6|
|SOURCE: U.S. Census Bureau (2004).|
4.7.1 Each state should publish a report on its voter education spending and activities.
4.7.2 States should engage in appropriate voter education efforts in coordination with local election authorities to assure that all citizens in their state have the information necessary to participate in the election process.
4.7.3 Each state should use its best efforts to instruct all high school students on voting rights and how to register to vote. In addition, civic education programs should be encouraged in the senior year of high school, as these have been demonstrated to increase voter participation by youth.
4.7.4 Local election authorities should mail written notices to voters in advance of an election advising the voter of the date and time of the election and the polling place where the voter can cast a ballot and encouraging the citizens to vote.The notice should also provide a phone number for the voter to contact the election authorities with any questions.
4.7.5 States should mail pamphlets to voters, and post the pamphlet material on their Web sites, to provide information about the candidates for statewide office and about ballot
initiatives and referenda.
4.7.6 The federal government should provide matching funds for the states to encourage civic and voter education and advertisements aimed to encourage people to vote.
^ Back to top
Because the integrity of the ballot is a hallmark of democracy, it is imperative that election officials guarantee eligible voters the opportunity to vote, but only once, and tabulate ballots in an accurate and fair manner.
While election fraud is difficult to measure, it occurs. The U.S. Department of Justice has launched more than 180 investigations into election fraud since October 2002. These investigations have resulted in charges for multiple voting, providing false information on their felon status, and other offenses against 89 individuals and in convictions of 52 individuals. The convictions related to a variety of election fraud offenses, from vote buying to submitting false voter registration information and voting-related offenses by non-citizens.54
In addition to the federal investigations, state attorneys general and local prosecutors handle cases of election fraud. Other cases are never pursued because of the difficulty in obtaining sufficient evidence for prosecution or because of the low priority given to election fraud cases. One district attorney, for example, explained that he did not pursue allegations of fraudulent voter registration because that is a victimless and nonviolent crime.55
Election fraud usually attracts public attention and comes under investigation only in close elections. Courts may only overturn an election result if there is proof that the number of irregular or fraudulent votes exceeded the margin of victory. When there is a wide margin, the losing candidate rarely presses for an investigation. Fraud in any degree and in any circumstance is subversive to the electoral process. The best way to maintain ballot integrity is to investigate all credible allegations of election fraud and otherwise prevent fraud before it can affect an election.
Investigation and prosecution of election fraud should include those acts committed by individuals, including election officials, poll workers, volunteers, challengers or other nonvoters associated with the administration of elections, and not just fraud by voters.
5.1.1 In July of even-numbered years, the U.S. Department of Justice should issue a public report on its investigations of election fraud.This report should specify the numbers of allegations made, matters investigated, cases prosecuted, and individuals convicted for various crimes. Each state’s attorney general and each local prosecutor should issue a similar report.
5.1.2 The U.S. Department of Justice’s Office of Public Integrity should increase its staff to investigate and prosecute election-related fraud.
5.1.3 In addition to the penalties set by the Voting Rights Act, it should be a federal felony for any individual, group of individuals, or organization to engage in any act of violence, property destruction (of more than $500 value), or threatened act of violence that is intended to deny any individual his or her lawful right to vote or to participate in a federal election.
5.1.4 To deter systemic efforts to deceive or intimidate voters, the Commission recommends federal legislation to prohibit any individual or group from deliberately providing the public with incorrect information about election procedures for the purpose of preventing voters from going to the polls.
Fraud occurs in several ways. Absentee ballots remain the largest source of potential voter fraud.56 A notorious recent case of absentee ballot fraud was Miami’s mayoral election of 1998, and in that case, the judge declared the election fraudulent and called for a new election. Absentee balloting is vulnerable to abuse in several ways: Blank ballots mailed to the wrong address or to large residential buildings might get intercepted. Citizens who vote at home, at nursing homes, at the workplace, or in church are more susceptible to pressure, overt and subtle, or to intimidation. Vote buying schemes are far more difficult to detect when citizens vote by mail. States therefore should reduce the risks of fraud and abuse in absentee voting by prohibiting “third-party” organizations, candidates, and political party activists from handling absentee ballots. States also should make sure that absentee ballots received by election officials before Election Day are kept secure until they are opened and counted.
Non-citizens have registered to vote in several recent elections. Following a disputed 1996 congressional election in California, the Committee on House Oversight found 784 invalid votes from individuals who had registered illegally. In 2000, random checks by the Honolulu city clerk’s office found about 200 registered voters who had admitted they were not U.S. citizens.57 In 2004, at least 35 foreign citizens applied for or received voter cards in Harris County, Texas, and non-citizens were found on the voter registration lists in Maryland as well.58
The growth of “third-party” (unofficial) voter registration drives in recent elections has led to a rise in reports of voter registration fraud. While media attention focused on reports of fraudulent voter registrations with the names of cartoon characters and dead people, officials in 10 states investigated accusations of voter registration fraud stemming from elections in 2004, and between October 2002 and July 2005, the U.S. prosecuted 19 people charged with voter registration fraud.59 Many of these were submitted by third-party organizations, often by individuals who were paid by the piece to register voters.
States should consider new legislation to minimize fraud in voter registration, particularly to prevent abuse by third-party organizations that pay for voter registration by the piece. Such legislation might direct election offices to check the identity of individuals registered through third-party voter registration drives and to track the voter registration forms.
HAVA requires citizens who register by mail to vote in a state for the first time to provide an ID when they register or when they vote. Some states have interpreted this requirement to apply only to voter registration forms sent to election offices by mail, not to forms delivered by third-party organizations. As a result, neither the identity nor the actual existence of applicants is verified. All citizens who register to vote with a mail-in form, whether that form is actually sent by mail or is instead hand-delivered, should comply with HAVA’s requirements or with stricter state requirements on voter ID, by providing proof of identity either with their registration application or when they appear at the polling station on Election Day. In this way, election offices will be obliged to verify the identity of every citizen who registers to vote, whether or not the registration occurs in person.
In addition, states should introduce measures to track voter registration forms that are handled by third-party organizations. By assigning a serial number to all forms, election officials will be able to track the forms. This, in turn, will help in any investigations and prosecutions and thus will serve to deter voter registration fraud.
Many states allow the representatives of candidates or political parties to challenge a person’s eligibility to register or vote or to challenge an inaccurate name on a voter roll. This practice of challenges may contribute to ballot integrity, but it can have the effect of intimidating eligible voters, preventing them from casting their ballot, or otherwise disrupting the voting process. New procedures are needed to protect voters from intimidating tactics while also offering opportunities to keep the registration rolls accurate, and to provide observers with meaningful opportunities to monitor the conduct of the election. States should define clear procedures for challenges, which should mainly be raised and resolved before the deadline for voter registration. After that, challengers will need to defend their late actions. On Election Day, they should direct their concerns to poll workers, not to voters directly, and should in no way interfere with the smooth operation of the polling station.
5.2.1 State and local jurisdictions should prohibit a person from handling absentee ballots other than the voter, an acknowledged family member, the U.S. Postal Service or other legitimate shipper, or election officials.The practice in some states of allowing candidates or party workers to pick up and deliver absentee ballots should be eliminated.
5.2.2 All states should consider passing legislation that attempts to minimize the fraud that has resulted from “payment by the piece” to anyone in exchange for their efforts in voter registration, absentee ballot, or signature collection.
5.2.3 States should not take actions that discourage legal voter registration or get-out-thevote activities or assistance, including assistance to voters who are not required to vote in person under federal law.
^ Back to top
To build confidence in the electoral process, it is important that elections be administered in a neutral and professional manner. Election officials, from county clerks and election board members to secretaries of state and U.S. Election Assistance Commission members, generally have shown great skill and dedication in administering elections in a fair and impartial manner. The institutions of election administration, however, are in need of improvement, so that they may instill greater public confidence in the election process and allow election officials to carry out their responsibilities more effectively (see Table 5 on page 52).
Elections are contests for power and, as such, it is natural that politics will influence every part of the contest, including the administration of elections. In recent years, some partisan election officials have played roles that have weakened public confidence in the electoral process. Many other partisan election officials have tried to execute their responsibilities in a neutral manner, but the fact that they are partisan sometimes raises suspicions that they might favor their own party. Most other democratic countries have found ways to insulate electoral administration from politics and partisanship by establishing truly autonomous, professional, and nonpartisan independent national election commissions that function almost like a fourth branch of government. The United States, too, must take steps to conduct its elections impartially both in practice and in appearance.
Impartial election administration, however, is not enough. Elections must also be administered effectively if they are to inspire public confidence. Long lines at polling stations, inadequately trained poll workers, and inconsistent or incorrect application of electoral procedures may have the effect of discouraging voter participation and may, on occasion, raise questions about bias in the way elections are conducted. While problems at polling stations usually reflect a shortage of trained poll workers or poor management of polling station operations, rather than an attempt to seek partisan advantage, the result is much the same. Such problems raise public suspicions or may provide grounds for the losing candidate to contest the result in a close election.
The intense partisanship and the close division of the American electorate, coupled with the Electoral College system, raise the possibility of another presidential election decided by a razor-thin margin in one or more battleground states. Although voting technology is improving, presidential elections are held in a decentralized system with a patchwork of inconsistent rules. In addition, in recent years, election challenges in the courts have proliferated.
Close elections, especially under these conditions, put a strain on any system of election administration, and public opinion demonstrates this. Significant segments of the American public have expressed concern about voter fraud, voter suppression, and the fairness of the election process in general.60 While substantially more Democrats than Republicans surveyed in national polls considered the 2004 presidential election unfair, 41 percent more Republicans than Democrats said the electoral process was unfair in Washington state’s 2004 gubernatorial election, which the Democratic candidate won by a very narrow margin.61 The losing side, not surprisingly, is unhappy with the election result, but what is new and dangerous in the United States is that the supporters of the losing side are beginning to believe that the process is unfair. And this is true of both parties.
At its base, the problem is a combustible mixture of partisan suspicion and irregularities born in part from a decentralized system of election administration with differing state laws determining voter registration and eligibility and whether a ballot is actually counted. The irregularities, by and large, stem from a lack of resources and inadequate training for election workers, particularly those who work just on Election Day. In other countries, such irregularities sometimes lead to street protests or violence. In the United States, up until now, we have been relatively fortunate that irregularities are addressed in court. The dramatic increase in election-related litigation in recent years, however, does not enhance the public’s perception of elections and may in fact weaken public confidence. The average number of election challenges per year has increased from 96 in the period of 1996 to 1999 to 254 in 2001 to 2004.62
Another major source of public mistrust of the election process is the perception of partisanship in actions taken by partisan election officials. In a majority of states, election administration comes under the authority of the secretary of state. In 2000 and 2004, both Republican and Democratic secretaries of state were accused of bias because of their discretionary decisions — such as how to interpret unclear provisions of HAVA. The issue is not one of personality or a particular political party because allegations and irregularities dogged officials from both parties. The issue is the institution and the perception of partiality that is unavoidable if the chief election officer is a statewide politician and the election is close, has irregularities, or is disputed. The perception of partiality is as important, if not more so, than the reality.
Bipartisan election administration has the advantage of allowing both parties to participate, but the flaws of such a system are evident in the experience of the Federal Election Commission (FEC). The FEC has often become deadlocked on key issues. In the cases when the FEC commissioners agree, they sometimes protect the two parties from enforcement rather than represent the public’s interest in regulating campaign finance.
NONPARTISAN ELECTION ADMINISTRATION. To minimize the chance of election meltdown and to build public trust in the electoral process, nonpartisan structures of election administration are very important, and election administrators should be neutral, professional, and impartial. At the federal level, the U.S. Election Assistance Commission should be reconstituted on a nonpartisan basis to exercise whatever powers are granted by law, and the EAC chairperson should serve as a national spokesperson, as the chief elections officer in Canada does, for improving the electoral process. States should consider transferring the authority for conducting elections from the secretary of state to a chief election officer, who would serve as a nonpartisan official.
States could select a nonpartisan chief elections officer by having the individual subject to approval by a super-majority of two-thirds of one or both chambers of the state legislature. The nominee should receive clear bipartisan support. This selection process is likely to yield a respected consensus candidate or, at least, a nonpartisan candidate.
The EAC, in its 18 months of operation, has managed to make its decisions by consensus. While this is a significant accomplishment for a bipartisan, four-member commission, it has come at a cost. The EAC has been slow to issue key guidance, and the guidance it has issued has often been vague. The process of forging consensus among the EAC’s commissioners appears to have slowed and watered down key decisions, particularly as they have come under pressure from their respective political parties. If the EAC were reconstituted as a nonpartisan commission, it would be better able to resist partisan political pressure and operate more efficiently and effectively.
To avoid the dangers of bipartisan stalemate, the EAC should be reconstituted as a fivemember commission, with a strong chairperson and nonpartisan members. This would be done initially by adding a fifth position to the EAC and making that position the chairperson, when the current chairperson’s term ends. The new EAC chairperson would be nonpartisan, nominated by the President, and confirmed by the U.S. Senate. Later, as the terms of other EAC commissioners expired, they would be replaced by nonpartisan commissioners, subject to Senate confirmation as well.
INDEPENDENCE AND AUTHORITY. For the positions of EAC commissioners and state chief elections officers to remain both nonpartisan and effective, they must be insulated from political pressure. This can be done by the terms of appointment and the lines of responsibility. The EAC commissioners and state chief elections officers should receive a long-term appointment, perhaps 10 years. The grounds for dismissal should be limited, similar to the rules for removal of a federal or state judge. The EAC should have the autonomy to oversee federal election laws that Congress directs it to implement and advise Congress and the President on needed improvements in election systems. State chief elections officers should have similar autonomy.
Under HAVA, the EAC distributes federal funds to the states, issues voluntary guidance on HAVA’s mandates, and serves as a clearinghouse for information on elections. In addition, it develops standards for voting equipment and undertakes research on elections.
The flaws identified in the electoral system described in this report were due in large part to a very decentralized system with voting standards implemented in different ways throughout the country. If HAVA is fully and effectively implemented, states should be able to retrieve authority to conduct elections from counties and impose a certain degree of uniformity.
In this report, we have proposed the kinds of reforms needed to improve significantly our electoral process. To implement those reforms, a new or invigorated institution like the EAC is needed to undertake the following tasks:
|TABLE 5: Types of Electoral Administration|
W O R L D R E G I O N
Asia & the
(percent of total)
|Government supervised by judges or others||6||2||6||14||28 (23%)|
|Independent electoral commission||25||19||12||19||75 (63%)|
|* The U.S. is included in this category.
SOURCE: Rafael López-Pintor. Electoral Management Bodies as Institutions of Governance (NY: United Nations Development Programme, Bureau for Development Policy, 2000).
These reforms, but particularly those that require connecting states, will not occur on their own. The EAC needs to have sufficient authority to assure effective and consistent implementation of these reforms, and to avoid repeating past problems, its guidance must be clear and compelling. A stronger EAC does not mean that the states will lose power in conducting elections. To the contrary, the authority of state election officials will grow with the creation of statewide voter databases, and their credibility will be enhanced by the new nonpartisan structure and professionalism.
CONFLICT-OF-INTEREST RULES. No matter what institutions are responsible for conducting elections, conflict-of-interest standards should be introduced for all federal, state, and local election officials, including some of the provisions in Colorado’s new election law and of the Code of Conduct prepared by the International Institute for Democracy and Electoral Assistance (IDEA).63 This Code of Conduct requires election administrators to avoid any activity, public or private, that might indicate support or even sympathy for a particular candidate, political party, or political tendency.
Election officials should be prohibited by federal and/or state laws from serving on any political campaign committee, making any public comments in support of a candidate, taking a public position on any ballot measure, soliciting campaign funds, or otherwise campaigning for or against a candidate for public office. A decision by a secretary of state to serve as co-chair of his or her party’s presidential election committee would clearly violate these standards.
6.1.1 To undertake the new responsibilities recommended by this report and to build confidence in the administration of elections, Congress and the states should reconstitute election management institutions on a nonpartisan basis to make them more independent and effective. U.S. Election Assistance Commission members and each state’s chief elections officer should be selected and be expected to act in a nonpartisan manner, and the institutions should have sufficient funding for research and training and to conduct the best elections possible.We believe the time has come to take politics as much as possible out of the institutions of election administration and to make these institutions nonpartisan.
6.1.2 Congress should approve legislation that would add a fifth member to the U.S. Election Assistance Commission, who would serve as the EAC’s chairperson and who would be nominated by the President based on capability, integrity, and nonpartisanship.This would permit the EAC to be viewed more as nonpartisan than bipartisan and would improve its ability to make decisions.That person would be subject to Senate confirmation and would serve a single term of ten years. Each subsequent vacancy to the EAC should be filled with a person judged to be nonpartisan so that after a suitable period, all the members, and thus the institution, might be viewed as above politics.
6.1.3 States should prohibit senior election officials from serving or assisting political campaigns in a partisan way, other than their own campaigns in states where they are elected.
6.1.4 States should take additional actions to build confidence in the administration of elections by making existing election bodies as nonpartisan as possible within the constraints of each state’s constitution. Among the ways this might be accomplished would be if the individuals who serve as the state’s chief elections officer were chosen based on their capability, integrity, and nonpartisanship.The state legislatures would need to confirm these individuals by a two-thirds majority of one or both houses.The nominee should receive clear bipartisan support.
6.1.5 Each state’s chief elections officer should, to the extent reasonably possible, ensure uniformity of voting procedures throughout the state, as with provisional ballots. Doing so will reduce the likelihood that elections are challenged in court.
For generations, civic-minded citizens, particularly seniors, have served as poll workers. The average age of poll workers is 72.64 Poll workers generally are paid minimum wages for a 15- hour day. Not surprisingly, recruitment has proven more and more difficult. For the 2004 election, the United States needed 2 million poll workers, but it fell short by 500,000.
Effective administration of elections requires that poll workers have the capability and training needed to carry out complex procedures correctly, the skills to handle increasingly sophisticated voting technology, the personality and skills to interact with a diversity of people in a calm and friendly manner, and the energy to complete a very long and hard day of work on Election Day. Poll workers must administer complex voting procedures, which are often changed with each election. These procedures include issuing provisional ballots, checking voter identification in accordance with state law, and correctly counting the votes after the polling station closes. Poll workers must also set up voting machines, instruct voters to use these machines, and provide helpful service to voters, including to voters with disabilities and non-English speakers.
A broad pool of potential recruits, drawn from all age groups, is needed to meet the demands made on today’s poll workers. To adequately staff polling stations, states and local jurisdictions must offer better pay, training, and recognition for poll workers and recruit more citizens who have full-time jobs or are students. Recruitment of teachers would serve to spread knowledge of the electoral process, while recruitment of students would educate future voters and attract individuals who may serve as poll workers for decades to come.
Local election authorities should also consider providing incentives for more rigorous training. Guilford County, North Carolina, for example, initiated a “Precinct Officials Certification” program in cooperation with the local community college. The program requires 18 hours of class and a final exam. While voluntary, more than 80 percent of Guilford County’s 636 permanent precinct officials completed the course. Certified officials receive an additional $35 per election in pay. Retention of officials has risen from roughly 75 percent to near 95 percent.
In addition, poll workers deserve greater recognition for their public service. States might establish a Poll Worker Appreciation Week and issue certificates to thank poll workers for their contribution to the democratic process.
Several states have passed laws to provide paid leave for state and local government workers who serve as poll workers on Election Day. A pilot program titled “Making Voting Popular” was implemented in 1998 in six counties surrounding the Kansas City metropolitan area to encourage employers to provide a paid “civic leave” day for employees who work as poll workers. Many states have introduced laws to encourage the recruitment of student poll workers. Partnered with experienced poll workers, student poll workers can learn about elections while contributing their technological skills.
It will be easier to recruit skilled poll workers if they are given flexibility in the terms of their service by working part of the day. Since a large proportion of voters arrive either at the beginning or the end of the day, it would make sense to hire more poll workers for those periods, although this is not now the case. Bringing poll workers in from other jurisdictions might also serve to provide partisan balance in jurisdictions where one party is dominant. Flexibility in the terms of service by poll workers is often restricted by state laws. Where this is the case, states should amend their laws to allow part-day shifts for poll workers on Election Day and to permit state residents to staff polling stations in a different jurisdiction.
In addition, states might consider a new practice of recruiting poll workers in the same way that citizens are selected for jury duty. This practice is used in Mexico, where citizens are selected randomly to perform what they consider a civic obligation. About five times as many poll workers as needed are trained in Mexico, so that only the most skilled and committed are selected to serve as poll workers on Election Day. The process of training so many citizens serves the additional purpose of educating the public in voting procedures. This practice both reflects and contributes to a broad civic commitment to democracy.
6.2.1 States and local jurisdictions should allocate sufficient funds to pay poll workers at a level that would attract more technologically sophisticated and competent workers. Part-time workers should also be recruited for the beginning and the end of Election Day. States should amend their laws to allow shifts for part of the day for poll workers on Election Day.
6.2.2 States and local jurisdictions should implement supplemental training and recognition programs for poll workers.
6.2.3 To increase the number and quality of poll workers, the government and nonprofit and private employers should encourage their workers to serve as poll workers on Election Day without any loss of compensation, vacation time or personal time off. Special efforts should be made to enlist teachers and students as poll workers.
6.2.4 Because some jurisdictions have large majorities of one party, which makes it hard to attract poll workers from other parties, local jurisdictions should allow poll workers from outside the jurisdiction.
6.2.5 States should consider legislation to allow the recruitment of citizens as poll workers as is done for jury duty.
A visible problem on Election Day 2004 was long lines. This should have been anticipated because there was a surge in new registrations and people expected a close election, particularly in “battleground states.” Still, too many polling stations were unprepared. While waiting until 4 a.m. to vote was an extreme case, too many polling stations experienced long lines at the beginning of the day when people went to work or at the day’s end when they returned. Fast-food chains hire extra workers at lunchtime, but it apparently did not occur to election officials to hire more workers at the times when most people vote. Long lines were hardly the only problem; many polling stations had shortages of provisional ballots, machines malfunctioned, and there were too many inadequately trained workers on duty. Although most states ban campaigning within a certain distance of a polling station, other states or counties permit it, though many voters find it distasteful if not intimidating.
Problems with polling station operations, such as long lines, were more pronounced in some places than in others.65 This at times gave rise to suspicions that the problems were due to discrimination or to partisan manipulation, when in fact the likely cause was a poor decision by election administrators. The U.S. Department of Justice’s investigation into the allocation of voting machines in Ohio, for example, found that problems were due to administrative miscalculations, not to discrimination.66
The 2004 elections highlighted the importance of providing enough voting machines to each polling place. While voter turnout can be difficult to predict, the ratio of voters per machine can be estimated. Texas, for example, has issued an administrative rule to estimate the number of machines needed per precinct at different rates of voter turnout.67
The impression many voters get of the electoral process is partially shaped by their experience at the polling station, and yet, not enough attention has been given to trying to make them “user-friendly.” Elementary questions, which most businesses study to become more efficient and responsive to their customers, are rarely asked, let alone answered by election officials. Questions like: How long does it normally take for a citizen to vote? Would citizens prefer to go to a neighborhood precinct, or to a larger, more service-oriented but more distant “voting center”? How many and what kinds of complaints and problems do polling stations hear in an average day? How do they respond, and are voters satisfied with the response? How many citizens find electronic machines useful, and how many find them formidable? By answering these fundamental questions, we might determine ways to provide efficient and courteous service at polling locations.
A simple way to compile useful information about problems voters face on Election Day would be to require that every voting station maintain a “log book” on Election Day to record all complaints from voters or observers. The log book would be signed by election observers at the end of the day to make sure that it has recorded all the complaints or problems. An analysis of the log books would help identify common problems and help design more efficient and responsive polling sites.
6.3.1 Polling stations should be made user-friendly. One way to do so would be to forbid any campaigning within a certain distance of a polling station.
6.3.2 Polling stations should be required to maintain a “log-book” on Election Day to record all complaints.The books should be signed by election officials and observers and analyzed for ways to improve the voting process.
6.3.3 Polling stations should be organized in a way that citizens would not have to wait long before voting, and officials should be informed and helpful.
Despite the wealth of expertise and literature on U.S. elections and voting behavior, little research focuses on the administration or conduct of elections. Until the 2000 election stirred interest in the subject, we had no information on how often votes went uncounted. Today, we still do not know how many people are unable to vote because their name is missing from the registration list or their identification was rejected at the polls. We also have no idea about the level of fraud or the accuracy and completeness of voter registration lists.
To effectively address the challenges facing our election systems, we need to understand better how elections are administered. The log books and public reports on investigations on election fraud, described above, can provide some good raw material. But we need more systematic research to expand knowledge and stimulate needed improvements in U.S. election systems. Moreover, beyond the reforms needed today, U.S. election systems will need to adapt in the future to new technology and to social changes.
The Center for Election Systems at Kennesaw State University in Georgia is the first university center established to study election systems and to assist election administration. With funding from the state government, this Center develops standards for voting technology used in Georgia and provides an array of other services, such as testing all election equipment, providing training, building databases, and designing ballots for many counties. The Center thus provides critical services to state election authorities and supports constant improvements in election systems. Since election laws and procedures vary significantly, each state should consider supporting university centers for the study of elections.
In addition to research on technology, university election centers could assist state governments on issues of election law, management, and civic and voter education. They could assemble experts from different disciplines to assist state governments in reviewing election laws, improving administrative procedures, strengthening election management, and developing programs and materials to train poll workers.
Comparative research is also needed on electoral systems in different states, and national studies should be conducted on different elements of election administration and causes of voter participation. These studies might address such questions as: What factors stimulate or depress participation in elections? How do voters adapt to the introduction of new voting technologies? And what are the costs of conducting elections? Research on these and a host of other questions is needed at the national, state, and local levels, with findings shared and efforts coordinated. Moreover, federal, state, and private foundation funds are needed to generate the research our election systems require to effectively inform decision-making, to monitor and advance best practices, and to measure implementation and enforcement.68
6.4.1 The Commission calls for continuing research on voting technology and election management so as to encourage continuous improvements in the electoral process.
Based on the limited available information, the cost of elections appears to vary significantly by state. Wyoming, for example, spent $2.15 per voter for the 2004 elections, while California spent $3.99 per voter.69 Information on the cost of elections is difficult to obtain, because both state and local authorities are involved in running elections, and local authorities often neglect to track what they spend on elections. At the county level, elections typically are run by the county clerk and recorder, who rarely keeps track of the staff time and office resources allocated to elections as opposed to other office responsibilities.
Election administration expenditures in the United States are on the low end of the range of what advanced democracies spend on elections. Among advanced democracies, expenditures on election administration range from lows of $2.62 in the United Kingdom and $3.07 in France for national legislative elections, through a midrange of $4.08 in Spain and $5.68 in Italy, to a high of $9.30 in Australia and $9.51 in Canada.70 While larger expenditures provide no guarantee of greater quality in election administration, they tend to reflect the priority given to election administration. The election systems of Australia and Canada are the most expensive but are also considered among the most effective and modern election systems in the world. Both local and state governments should track and report the cost of elections per registered voter. This data would be very important in offering comparisons on alternative and convenience voting.
6.5.1 As elections are a bedrock of our nation’s democracy, they should receive high priority in the allocation of government resources at all levels. Local jurisdictions, states, and the Congress should treat elections as a high priority in their budgets.
6.5.2 Both local and state governments should track and report the cost of elections per registered voter.
^ Back to top
The media’s role in elections is of great consequence. Effective media coverage contributes substantially to the electoral process by informing citizens about the choices they face in the elections and about the election results. In contrast, irresponsible media coverage weakens the quality of election campaigns and the public’s confidence in the electoral process.
More than $1.6 billion was spent on television ads in 2004 by candidates, parties, and independent groups.71 This was a record for any campaign year and double the amount spent in the 2000 presidential election.
The pressure to raise money to pay for TV ads has tilted the competitive playing field in favor of well-financed candidates and has created a barrier to entry in politics. Moreover, TV ads tend to reduce political discourse to its least attractive elements—campaign spots are often superficial and negative. This has a significant impact on the quality of campaigns, as television is the primary source of campaign information for about half of all Americans.72
Broadcasters receive free licenses to operate on our publicly owned airwaves in exchange for a pledge to serve the public interest. At the heart of this public interest obligation is the need to inform the public about the critical issues that will be decided in elections.
In 1998, a White House advisory panel recommended that broadcasters voluntarily air at least five minutes of candidate discourse every night in the month preceding elections. The goal of this “5/30 standard” was to give television viewers a chance to see candidates in nightly forums that are more substantive than the political ads that flood the airwaves in the final weeks of election campaigns. National networks were encouraged to broadcast a nightly mix of interviews, mini-debates, and issue statements by presidential candidates, and local stations were asked to do the same for candidates in federal, state, and local races. Complete editorial control over the forums for candidate discourse was, of course, left to the national networks and local stations, which would decide what campaigns to cover, what formats to use, and when to broadcast the forums.
In 2000, about 103 television stations pledged to provide at least five minutes of campaign coverage every night in the final month of the election campaign, yet they often fell short of the 5/30 standard. Local news broadcasts of these 5/30 stations provided coverage, on average, of only two minutes and 17 seconds per night of candidate discourse.73 On the thousand-plus stations that did not pledge to meet the 5/30 standard, coverage of candidate discourse was minimal.
During the 2004 campaign, substantive coverage of candidate discourse was still modest:74
Notwithstanding the dramatic expansion of news available on cable television, broadcasters can and should do more to improve their coverage of campaign issues. Some propose to require broadcasters to provide free air time to candidates, but others are concerned that it might lead toward public financing of campaigns or violate the First Amendment.
7.1.1 The Commission encourages national networks and local TV stations to provide at least five minutes of candidate discourse every night in the month leading up to elections.
7.1.2 The Commission encourages broadcasters to continue to offer candidates short segments of air time to make issue statements, answer questions, or engage in mini-debates.
7.1.3 Many members of the Commission support the idea that legislation should be passed to require broadcasters to give a reasonable amount of free air time to political candidates, along the lines of the provisions of the Our Democracy, Our Airwaves Act of 2003 (which was introduced as S.1497 in the 108th Congress).
For decades, early projections of presidential election results have diminished participation in the electoral process. Projections of Lyndon Johnson’s victory in 1964 came well before the polls closed in the West. The same occurred in 1972 and in 1980. In all of these cases, candidates further down the ballot felt the effect. In 1980, the estimated voter turnout was about 12 percent lower among those who had heard the projections and not yet voted as compared with those who had not heard the projections.75
On Election Night in 2000, the major television news organizations — ABC, CBS, NBC, CNN, and Fox — made a series of dramatic journalistic mistakes. While polls were still open in Florida’s panhandle, they projected that Vice President Gore had won the state. They later reversed their projection and predicted that Governor Bush would win Florida and, with it, the presidency. Gore moved to concede the election, beginning with a call to Bush. Gore later withdrew his concession, and the news organizations had to retract their projection of Bush’s victory. The first set of mistakes may have influenced voters in Florida and in other states where the polls were still open. The second set of mistakes irretrievably influenced public perceptions of the apparent victor in the election, which then affected the subsequent controversy over the outcome in Florida.
Having made these mistakes in 2000, most television news organizations were cautious about projecting presidential election results in 2004. This caution is worth repeating in future elections and should become a standard media practice.
The Carter-Ford Commission was highly critical of the practice of declaring a projected winner in a presidential election before all polls close in the contiguous 48 states of the United States. In the Commission’s view, this practice discourages voters by signaling that the election is over even before some people vote.
Voluntary restraint by major media organizations is a realistic option. National news networks in the last several presidential elections have voluntarily refrained from calling the projected presidential winner in the Eastern Standard Time zone until after 7:00 p.m. (EST). In addition, as a result of the mistakes they made in 2000, the networks have now agreed to refrain from calling the projected presidential winner in states with two time zones until all of the polls across the state have closed.
Media organizations should exercise similar restraint in their release of exit poll data. The Carter-Ford Commission noted the mounting body of evidence that documents the unreliability of exit polls. In 2000, exit polls conflicted with the actual election results in many states — and in five specific instances by as much as 7 percent to 16 percent. Network news organization officials acknowledged that exit polls have become more fallible over the years as more and more voters have refused to take part. In 2000, only about half of the voters asked to participate in exit polls agreed to do so, and only 20 percent of absentee and early voters agreed to participate in telephone “exit” poll interviews. That response rate is too low to assure reliability in exit polls.
Despite the effort made to improve exit polls for the 2004 presidential election, they were well off the mark and misled some Americans about the election’s outcome. By now it should be abundantly clear that exit polls do not reliably predict election results. While exit polls can serve a useful purpose after Election Day in providing data on the composition and preferences of the electorate, they lack credibility in projecting election results, and they reflect poorly on the news organizations that release them prematurely. This ought to give news organizations sufficient reason to abandon the practice of releasing exit poll data before elections have been decided.
Government cannot prohibit news organizations from irresponsible political reporting, and efforts to legislate a delay in the announcement of projected election results are problematic. Voluntary restraint on the part of news organizations offers the best recourse. By exercising voluntary restraint, news organizations will enhance their credibility and better serve the American people by encouraging participation and public confidence in elections.
7.2.1 News organizations should voluntarily refrain from projecting any presidential election results in any state until all of the polls have closed in the 48 contiguous states.
7.2.2 News organizations should voluntarily agree to delay the release of any exit poll data until the election has been decided.
^ Back to top
In too many states, election laws and practices do not allow independent observers to be present during crucial parts of the process, such as the testing of voting equipment or the transmission of results. In others, only certified representatives of candidates or political parties may observe. This limits transparency and public confidence in the election process. Above all, elections take place for the American people, rather than for candidates and political parties. Interested citizens, including those not affiliated with any candidate or party, should be able to observe the entire election process, although limits might be needed depending on the size of the group.
Although the United States insists on full access by its election observers to the elections of other countries, foreign observers are denied or granted only selective access to U.S. elections. Observers from the Organization for Security and Cooperation in Europe (OSCE), who were invited to the United States in 2004, were not granted access to polling stations in some states, and in other states, their access was limited to a few designated polling stations. Only one of our 50 states (Missouri) allows unfettered access to polling stations by international observers. The election laws of the other 49 states either lack any reference to international observers or fail to include international observers in the statutory categories of persons permitted to enter polling places.
To fulfill U.S. commitments to the OSCE “Copenhagen Declaration” on International Standards of Elections, accredited international observers should be given unrestricted access to U.S. elections. Such accreditation should be provided to reputable organizations which have experience in election observation and which operate in accordance with a recognized code of conduct. The National Association of Secretaries of State has encouraged state legislatures to make any necessary changes to state law to allow for international observers.76
8.1.1 All legitimate domestic and international election observers should be granted unrestricted access to the election process, provided that they accept election rules, do not interfere with the electoral process, and respect the secrecy of the ballot. Such observers should apply for accreditation, which should allow them to visit any polling station in any state and to view all parts of the election process, including the testing of voting equipment, the processing of absentee ballots, and the vote count. States that limit election observation only to representatives of candidates and political parties should amend their election laws to explicitly permit accreditation of independent and international election observers.
^ Back to top
The presidential primary system is organized in a way that encourages candidates to start their campaigns too early, spend too much money, and allow as few as eight percent of the voters to choose the nominees. The Commission believes that the scheduling of the presidential primary needs to be changed to allow a wider and more deliberate national debate.
In 2000, the presidential primaries were effectively over by March 9, when John McCain ended his bid for the Republican nomination and Bill Bradley left the race for the Democratic nomination. This was less than seven weeks after the Iowa caucus. In 2004, the presidential primary process was equally compressed. Less than 8 percent of the eligible electorate in 2004 cast ballots before the presidential nomination process was effectively over.
The presidential primary schedule has become increasingly front-loaded. While 8 states held presidential primaries by the end of March in 1984, 28 states held their primaries by March in 2004. The schedule continues to tighten, as six states have moved up the date of their presidential primary to February or early March while eight states have decided to cancel their presidential primary.77
Because the races for the presidential nominations in recent elections have generally concluded byMarch, most Americans have no say in the selection of presidential nominees, and intense media and public scrutiny of candidates is limited to about 10 weeks. Moreover, candidates must launch their presidential bids many months before the official campaign begins, so that they can raise the $25 to $50 million needed to compete.
The presidential primary schedule therefore is in need of a comprehensive overhaul. A new system should aim to expand participation in the process of choosing the party nominees for president and to give voters the chance to closely evaluate the presidential candidates over a three- to four-month period. Improvements in the process of selecting presidential nominees might also aim to provide opportunities for late entrants to the presidential race and to shift some emphasis from Iowa and New Hampshire to states that more fully reflect the diversity of America.
Most members of the Commission accept that the first two states should remain Iowa and New Hampshire because they test the candidates by genuine “retail,” door-to-door campaigning. A few other members of the Commission would replace those states with others that are more representative of America’s diversity, and would especially recommend a change from Iowa because it chooses the candidate by a public caucus rather than a secret ballot, the prerequisite of a democratic election.
While the presidential primary schedule is best left to the political parties to decide, efforts in recent years by political parties have failed to overhaul the presidential primary schedule. If political parties do not make these changes by 2008, Congress should legislate the change.
9.1.1 We recommend that the Chairs and National Committees of the political parties and Congress make the presidential primary schedule more orderly and rational and allow more people to participate.We endorse the proposal of the National Association of Secretaries of State to create four regional primaries, after the Iowa caucus and the New Hampshire primary, held at one-month intervals from March to June.The regions would rotate their position on the calendar every four years.
As the nation saw in 2000, a great deal of bitterness can arise when the outcome of a close presidential election turns on the interpretation of ambiguous laws. Had the U.S. Supreme Court not resolved the principal controversy in 2000, the dispute would have moved to Congress pursuant to Article II and the Twelfth Amendment. Unfortunately, the relevant provisions of the Constitution are vague or ambiguous in important respects, and the implementing legislation adopted by Congress over a century ago is not a model of clarity and consistency. If Congress is called upon to resolve a close election in the future, as could well happen, the uncertain meaning of these legal provisions is likely to lead to a venomous partisan spectacle that may make the 2000 election look tame by comparison.
After the debacle following the election of 1876, Congress spent more than a decade fashioning rules and procedures that it hoped would allow future disputes to be settled by preexisting rules. Those rules and procedures have remained on the books essentially unchanged since that time. The core provision (3 U.S.C. § 5) invites the states to establish appropriate dispute-resolution mechanisms by promising that Congress will give conclusive effect to the states’ own resolution of controversies if the mechanism was established before the election and if the disputes are resolved at least six days before the electoral college meets. This “safe-harbor” provision appropriately seeks to prevent Congress itself from having to resolve election disputes involving the presidency, and every state should take steps to ensure that its election statutes qualify the state for favorable treatment under the safe-harbor provision.
Unfortunately, even if all the states take this step, disputes requiring Congress to ascertain the meaning of unclear federal rules could still arise. Although it may not be possible to eliminate all possible sources of dispute, significant steps could be taken to improve the clarity and consistency of the relevant body of federal rules, and Congress should undertake to do so before the next presidential election.
9.2.1 Congress should clarify and modernize the rules and procedures applicable to carrying out its constitutional responsibilities in counting presidential electoral votes, and should specifically examine the deadlines.
9.2.2 States should certify their presidential election results before the “safe harbor” date. Also, every state should take steps, including the enactment of new statutes if necessary, to ensure that its resolution of election disputes will be given conclusive effect by Congress under 3 U.S.C. § 5.
^ Back to top
Building confidence in U.S. elections is central to our nation’s democracy. The vigor of our democracy depends on an active and engaged citizenry who believe that their votes matter and are counted accurately. The reforms needed to keep our electoral system healthy are an inexpensive investment in the stability and progress of our country.
As a nation, we need to pursue the vision of a society where most Americans see their votes as both a right and a privilege, where they cast their votes in a way that leaves them proud of themselves as citizens and of democracy in the United States. Ours should be a society where registering to vote is convenient, voting is efficient and pleasant, voting machines work properly, fraud is minimized, and disputes are handled fairly and expeditiously.
This report represents a comprehensive proposal for accomplishing those goals and modernizing our electoral system. We have sought to transcend partisan divides with recommendations that will both assure the integrity of the system and widen access. No doubt, there will be some who prefer some recommendations and others who prefer other proposals, but we hope that all will recognize, as we do, that the best way to improve our electoral system is to accept the validity of both sets of concerns.
The five pillars of our proposal represent an innovative and comprehensive approach. They break new ground in the following ways:
First, we propose a universal, state-based, top-down, interactive, and interoperable registration list that will, if implemented successfully, eliminate the vast majority of complaints currently leveled against the election system. States will retain control over their registration lists, but a distributed database offers a way to remove interstate duplicates and maintain an up-to-date, fully accurate registration list for the nation.
Second, we propose that all states require a valid photo ID card, which would be a slightly modified REAL ID or a photo ID that is based on an EAC-template (which is equivalent to the REAL ID without the drivers license). However, instead of allowing the ID to be a new barrier to voting, we propose using it to enfranchise new and more voters than ever before. The states would play a much more affirmative role of reaching out to the underserved communities by providing them more offices, including mobile ones, to register them and provide photo IDs free of charge. In addition, we offer procedural and institutional safeguards to make sure that the card is not abused and that voters will not be disenfranchised because of the need for an ID.
Third, we propose measures that will increase voting participation by connecting registration and the ID process, making voting more convenient, diminishing irregularities, and offering more information on voting.
Fourth, we propose ways to give confidence to voters that use the new electronic voting machines to ensure that their vote will be recorded accurately and there will be an auditable backup on paper (with the understanding that alternative technologies may be available in the future). Our proposals also aim to make sure that people with disabilities have full access to voting and the opportunity to do so privately and independently like other voters.
Finally, we recommend a restructuring of the system by which elections have been administered in our country. We propose that the Election Assistance Commission and state election management bodies be reconstituted on a nonpartisan basis to become more professional, independent, and effective.
Election reform is neither easy nor inexpensive. Nor can we succeed if we think of providing funds on a one-time basis. We need to view the administration of elections as a continuing challenge for the entire government, and one that requires the highest priority of our citizens and our government.
For more than two centuries, our country has taught the world about the significance of democracy, but more recently, we have evinced a reluctance to learn from others. Typical of this gap is that we insist other countries open their elections to international observers, but our states close their doors or set unfair restrictions on election observing. We recommend changing that provision and also building on the innovations of the new democracies by establishing new election management bodies that are independent, nonpartisan, and effective with a set of procedures that would make American democracy, once again, the model for the world.
The new electoral edifice that we recommend is built on the five pillars of reforms. Democrats, Republicans, and Independents may differ on which of these pillars are the most important, but we have come to understand that all are needed to improve our electoral system. Indeed, we believe that the structure is greater than the sum of its pillars. Substantively, the system’s integrity is strengthened by the increased access of its citizens, and voter confidence is raised by accuracy and security of new technology and enforcement of election laws. And the political support necessary to implement these reforms is more likely to materialize if all the pillars are viewed as part of an entire approach. If adequately funded and implemented, this new approach will move America down the path of transforming the vision of a model democracy into reality.
^ Back to top
The Commission’s recommendations are estimated to cost $1.35 billion to implement. This estimate is the sum of the cost of making state voter databases interoperable and upgrading voting machines to make them both accessible and transparent.
The total cost for making voter databases interoperable is estimated at $287 million. This cost breaks down as follows:
The total cost for upgrading voting machines, to make them both accessible and transparent, is estimated at $1.06 billion. This is the amount needed, in addition to the HAVA funds already obligated, to replace remaining punch card and lever machines with direct recording electronic (DRE) systems or with optical scan systems with a computerassisted marking device for blind and visually impaired voters, to retrofit DREs with a voter-verifiable paper audit trail, and to add a ballot marking device for blind voters to existing optical scan systems. The estimates are based on current distributions of various voting machines and on current costs for DREs, voter-verifiable paper audit trails, and ballot-marking devices for optical scan systems.
The Commission recommends that Congress provide $1.35 billion in funding over a twoyear period, so that voter databases will be made interoperable and voting machine upgrades will be completed before the 2008 elections.
^ Back to top
1 Adam Nagourney and Janet Elder, “Late Poll Still Shows Sharp Split in U.S. Vote,” International Herald Tribune, November 1, 2004; and Dan Eggen, “Justice Department Triples Election Monitors: More than 1,000 Head to Polls,” The Washington Post, October 29, 2004, p. A6.
2 The Pew Research Center for the People and the Press, “Voters Liked Campaign 2004, But
Too Much ‘Mud-Slinging’,” November 11, 2004, available at
3 Milwaukee Police Department, Milwaukee County District Attorney’s Office, Federal
Bureau of Investigation, and United States Attorney’s Office Task Force, Preliminary
Findings of Joint Task Force Investigating Possible Election Fraud. May 10, 2005. Available at
4 “Dead voters on rolls,” Chicago Tribune, December 4, 2004.
5 The following democracies constitute some of the nearly 100 countries that utilize a
national ID system: Belgium, Cost Rica, Germany, India, Italy, the Netherlands, Portugal,
South Africa, and Spain. See Privacy.org, “Identity Cards: FAQ,” August 24, 1996, available
6 Jason P. Schacter, “Geographical Mobility: 2002 to 2003.” Current Population Reports. US Census Bureau (March 2004). Available at: http://www.census.gov/prod/2004pubs/p20- 549.pdf.
7 In addition to the 38 states with top-down voter registration systems, 6 states are developing
bottom-up systems, 2 will use systems with both top-down and bottom-up elements, and 3
have yet to finalize their plans. North Dakota does not require voter registration. See
Electionline.org, Assorted Rolls: Statewide Voter Registration Databases Under HAVA, June
2005, p. 3, available at
8 “Exposed: Scandal of double voters,” New York Daily News, August 21, 2004 and “Double votes taint Florida, records show,” Orlando Sentinel, October 23, 2004.
9 “Report: As many as 60,000 people file to vote in both Carolinas,” Associated Press, October 24, 2004.
10 “Exposed: Scandal of Double Voters,” New York Daily News, August 21, 2004.
11 The introduction of electronic transaction standards would also facilitate cross-state exchanges of voter data, see R. Michael Alvarez and Thad E. Hall, “The Next Big Election Challenge: Developing Electronic Data Transaction Standards for Election Administration,” Caltech/MIT Voting Technology Project, July 2005, pp. 19-21.
12 “Overview of States Driver’s License Requirements”, National Immigration Law Center,
July 12, 2005, available at
13 Except for Vermont, all states require a Social Security Number for a driver’s license, at least from people who were assigned a Social Security Number or are eligible for one.
14 Voters should also have the opportunity to check their registration over the phone, via a toll-free number, or in person at the elections office.
15 Electionline.org, Solution or Problem? Provisional Ballots in 2004, April 2005, p. 2, available
16 Ibid, p.5.
17 In states with unified databases, provisional ballots constituted .85 percent of the total ballots cast whereas in the states without unified databases, provisional ballots constituted 1.76 percent of the total. See Electionline.org, Solution or Problem? Provisional Ballots in 2004,Washington, D.C., April 2005.
18 Testimony before the Commission by Ken Smukler, President of Info Voter Technologies, on June 30, 2005.
19 Details were provided in Section 1.1.
20 ID is required of all voters in 22 states and of all first-time voters in another two states,
according to Electionline.org,
21 Provided by Electionline.org,
22 A comparison of driver’s license records and census data for 2003 suggests that about 88
percent of Americans aged 18 and over have a driver’s license, see U.S. Department of
Transportation, Federal Highway Administration, Licensed Total Drivers, By Age, 2003,
Table DL-22, Oct. 2004, at
23 U.S. Government Accountability Office, Elections: Additional Data Could Help State and Local Elections Officials Maintain Accurate Voter Registration Lists, GAO-05-478, June 2005, pp. 13-29.
24 U.S. Election Assistance Commission, The Impact of the National Voter Registration Act, 2003–2004, June 30, 2005, pp. 16 and 20.
25 Data on voter registration in Alaska is contained in U.S. Election Assistance Commission, “The Impact of the National Voter Registration Act of 1993 on the Administration of Elections for Federal Office: 2003-2004,” Table 1: Registration History. Other examples include 34 of the 82 counties in Mississippi and the City of East St. Louis, see Emily W. Pettus, “Secretary of state seeks proposals on statewide voter roll,” Associated Press, September 1, 2004, and Mike Fitzgerald, “Dual registration: a recipe for fraud?” Belleview News-Democrat, November 28, 2004.
26 For example, see Australian National Audit Office, Integrity of the Electoral Roll, April 2002:
27 The residual vote rates fell by 0.79 percent in counties where lever machines were replaced by direct recording electronic (DRE) machines and by 1.46 percent in counties where punch cards were replaced by DREs, according to Charles Stewart, Residual Vote in the 2004 Election, Caltech/MIT Voting Technology Project Working Paper, February 2005, Table 2.
28 Election Data Services,
29 Dan Keating, “Lost Votes in N.M. a Cautionary Tale,” Washington Post, August 22, 2004, and “Nearly 40 votes may have been lost in Palm Beach County,” Associated Press, November 2, 2004.
31 Ted Selker, “Processes Can Improve Electronic Voting,” Caltech/MIT Voting Technology
Project, October 2004, available at
32 Manual audits of voting machines are required in Colorado, Connecticut, Hawaii, Illinois,
Minnesota, New Mexico, New York, North Carolina, Washington, and West Virginia,
according to Verified Voting Foundation, “Manual Audit Requirement,” August 18, 2005,
33 Ted Selker and Jon Goler, “Security Vulnerabilities and Problems with VVPT,”
Caltech/MIT Voting Technology Project, April 2004, available at
34/sup> “Voting Machine Fails Inspection,” CNETNews.com, July 23, 2003 and “New Security Woes for E-Vote Firm,” WiredNews.com, August 7, 2003.
35 In California’s field test, about one in ten machines malfunctioned, see “Voting Machines Touch and Go,” Associated Press, July 30, 2005.
36 Internet Policy Institute, Report of the National Workshop on Internet Voting: Issues and
Research Agenda, March 2001. Available at
37 Curtis Gans, “Making it Easier Doesn’t Work: No Excuse Absentee and Early Voting Hurt Voter Turnout,” Center for the Study of the American Electorate, September 13, 2004, available at http://www.american.edu/ia/cfer/research/csae_09132004.pdf.
38 Testimony before the Commission by Robert Stein, Dean of Social Sciences at Rice University, on June 30, 2005.
39 Balancing Access and Integrity: The Report of the Century Foundation working Group on State Implementation of Election Reform (N.Y. the Century Foundation Press, 2005), pp. 25-26.
40 Curtis Gans, “Making it Easier Doesn’t Work: No Excuse Absentee and Early Voting Hurt Voter Turnout,” Center for the Study of the American Electorate, September 13, 2004, available at http://www.american.edu/ia/cfer/research/csae_09132004.pdf.
41 Superior Court of the State of Washington for Chelan County, Final Judgment Dismissing Election Contest with Prejudice and Confirming Certification of Election of Christine Gregoire, Court Decision No. 05-2-00027-3, June 6, 2005.
42 United States General Accounting Office. “Elections: Issues Affecting Military and Overseas
Absentee Voters,” May 2001, available at:
43 National Defense Committee, Military and Overseas Absentee Voting in the 2004 Presidential
Election, March 30, 2005, available at
44 David Jefferson, Aviel D. Rubin, Barbara Simons, and David Wagner, A Security Analysis of the Secure Electronic Registration and Voting Experiment, January 20, 2004,
45 Information provided to the Commission by the Federal Voting Assistance Program.
46 Testimony before the Commission by James Dickson, Vice President at the American Association of People with Disabilities, on April 18, 2005.
50 Alabama, Arizona, Delaware, Maryland, Mississippi, Nebraska, Nevada, Tennessee,
Washington, and Wyoming have a permanent ban on voting by certain categories of exfelons,
according to the Sentencing Project,
52 Karl T. Kurtz, Alan Rosenthal, and Cliff Zukin, Citizenship: A Challenge for All Generations, National Conference of State Legislatures, September 2003, available at
53 Campaign for the Civic Mission of Schools and Alliance for Representative Democracy, “From Classroom to Citizen: American Attitudes on Civic Education,” December 2004, available at
54 U.S. Department of Justice press release, “Department of Justice to Hold Ballot Access and Voting Integrity Symposium,” August 2, 2005.
55 U.S. Government Accountability Office, Elections: Additional Data Could Help State and Local Elections Officials Maintain Accurate Voter Registration Lists, GAO-05-478, June 2005, pp. 59-60.
56 Balancing Access and Integrity: The Report of the Century Foundation working Group on State Implementation of Election Reform (N.Y. the Century Foundation Press, 2005), pp. 67-69.
57 John Fund, Stealing Elections: How Voter Fraud Threatens Our Democracy (San Francisco: Encounter Books, 2004), p. 103.
58 Joe Stinebaker, “Loophole lets foreigners illegally vote,” Houston Chronicle, January 16, 2005, and Robert Redding, “Purging illegal aliens from voter rolls not easy: Maryland thwarted in tries so far,” Washington Times, August 23, 2004.
59 Susan Greene and Karen E. Crummy, “Vote Fraud Probed In State,” Denver Post, March 24, 2005; Brendan Farrington, “Fla. Officials Asked To Probe Vote Fraud,” Associated Press, October 7, 2004; Dawson Bell, “Campaign Workers Suspected Of Fraud,” Detroit Free Press, September 23, 2004; “Man Pleads Guilty In Voter Registration Scam,” Associated Press, December 7, 2004; Robert Patrick, “Jury Finds Montgomery Guilty In Vote Fraud Case,” St. Louis Post-Dispatch, February 11, 2005; Nevada Secretary Of State, “Alleged Vote Fraud Investigations Ongoing,” Press Release, October 28, 2004; Dan McKay, “Election ‘Mischief’ Under Scrutiny,” Albuquerque Journal, September 10, 2004; “Voter Registration Investigation One Of Largest In Recent Years,” Associated Press, September 23, 2004; Greg J. Borowski, “Inquiry Finds Evidence Of Fraud In Election,” Milwaukee Journal Sentinel, May 11, 2005; U.S. Department of Justice, Criminal Division, Public Integrity Section, Election Fraud Prosecutions and Convictions: Ballot Access & Voting Integrity Initiative, October 2002 - July, 2005.
60 A Rasmussen Reports poll just before the November 2004 elections showed that 58 percent of American voters believed there was “a lot” or “some” fraud in U.S. elections, and in a post-election NBC News/Wall Street Journal poll, more than a quarter of Americans worried that the vote count for president in 2004 was unfair, quoted in Rick Hasen, “Beyond the Margin of Litigation: Reforming Election Administration to Avoid Electoral Meltdown,” Paper prepared for American Political Science Association meeting, September 1, 2005, pp. 7-8, available at
61 Ibid, p. 9.
62 Ibid, p. 29.
63 International IDEA, Code of Conduct for the Ethical and Professional Administration of
64 United States Election Assistance Commission, Background on the Help America Vote College
Poll Worker Program.
65 The Voting Rights Institute, Democracy at Risk: the 2004 Election in Ohio (Washington, D.C.: Democratic National Committee, 2005).
66 U.S. Department of Justice’s investigations in Franklin County and in Knox County, Ohio found no evidence that the allocation of voting machines was conducted in a discriminatory manner, see
67 Rule §81.125 of Texas Administrative Code, available at
68 A strong example of funding for elections research is the $7.5 million awarded by the National Science Foundation on August 15, 2005 for a collaborative project of six institutions to study the reliability, security, transparency, and auditability of voting systems.
69 California Secretary of State Historical Close Of Registration Statistics: Presidential General Elections, May 2004, available at
70 IFES, Cost of Registration and Elections (CORE) for election costs in Australia and Spain; Elections Canada,
71 Alliance for Better Campaigns,
72 Fox New/Opinion Dynamics poll, March 25, 2004,
73 Analysis by the Norman Lear Center at the Annenberg School for Communication of the University of Southern California,
74 Alliance for Better Campaigns,
75 National Commission on Federal Election Reform, To Assure Pride and Confidence in the Electoral Process, August 2001, p. 63.
76 National Association of Secretaries of State, “International Election Protocol Resolution,” and supporting language, July 24, 2005, available at
77 Six states passed measures to move forward the date of their presidential primaries and eight states passed measures to cancel their presidential primary for 2004, see
78 Estimate is based on the average amounts other states are currently spending to build topdown voter registration systems and excludes HAVA funds that have already been disbursed for this purpose see Electiononline.org, Assorted Rolls: Statewide Voter Registration Databases Under HAVA,
79 Figure includes both the cost to upgrade existing state databases to make them interoperable in real time and the cost to build a voter registration distributed database linked to the individual state servers. The former ($48 million) is based on the average cost to make existing state driver’s license databases interoperable with each other as determined by the Congressional Budget Office, see “H.R. 418: REAL ID Act of 2005,” Congressional Budget Office
80 The cost to collect Social Security numbers is tantamount to registering voters. The Office of the Chief Electoral Officer of Canada calculates the cost to registering 19.6 million voters in the 1997 national elections at approximately $18 million. This produces a statistic of $0.92 to register each person, see Voter Turnout, electionguide.org,
81 The cost per card is estimated at $5. This figure includes approximate administrative, infrastructure, and issuance costs, see Stephen Moore, “Congressional testimony before the U.S. House of Representatives Subcommittee on Immigration and Claims, Judiciary Committee,” May 13, 1997, available at
82 The estimated costs for the various voting machines are as follows: Direct Recording Electronic with a Voter-Verified Paper Audit Trail (DRE/VVPAT)—$4,000; retrofitting a DRE machine with a VVPAT—$1,000; optical scanner (OS)—$5,000; and ballot marking device for an optical scan system—$4,500. Machine cost data is collected from many
sources, including: Verifiedvoting.org, “Appendix 4: Cost Comparison of Alternative Solutions,”
^ Back to top
^ Back to top
All of the Commission Members are signatories of the report. Some have submitted additional or dissenting statements, which they were asked to limit to 250 words.
For alternative views and additional comments on the Commission’s report, see our Web page at www.american.edu/ia/cfer/comments.
Kay Coles James
I strongly support the recommendation that states adopt uniform procedures for determining the validity of provisional ballots, and I join a majority of members who support counting provisional ballots when they are cast in the wrong precinct where multiple precincts vote at a single polling place.
However, out-of-precinct voting, in which a voter uses a provisional ballot to cast a ballot in the incorrect precinct, raises four substantial problems: (1) The voter is denied opportunity to vote for all candidates and issues or else casts a vote in a race in which the voter is not qualified to vote. (2) Election officials will not be able to anticipate the proper number of voters appearing at any given polling place and will not be able to allocate resources properly among the various polling places with the result that voters will face long lines and shortage of voting supplies. (3) The post-election evaluation of provisional ballots cast in the wrong polling place is time-consuming, error prone, subject to manipulation, undermines the secrecy of the ballot and will delay the outcome of the election. (4) It is settled law that HAVA does not mandate out-of-precinct voting.
The fact that many members of the Commission support limited out-of-precinct voting should not be understood as this Commission is recommending out-of-precinct voting because a substantial number of Commission members oppose it.
See Daschle, et. al. below for an alternative view of this recommendation.
Tom Daschle joined by Spencer Overton and Raul Yzaguirre
The goals of ballot access and integrity are not mutually exclusive, and the ultimate test of the Commission’s success will be whether voters from diverse backgrounds view its recommendations in their totality as providing them with a fair opportunity to participate in their democracy. Most of the recommendations in this report, such as the recommendation for a voter verified paper audit trail, meet that standard, but others do not. For voters who have traditionally faced barriers to voting – racial and ethnic minorities, Native Americans, the disabled and language minorities, the indigent and the elderly – these recommendations appear to be more about ballot security than access to the ballot.
The call for States to use the new REAL ID driver’s license for voter identification at the polls is the most troublesome recommendation in the Report. While this statement identifies some of its problems, unfortunately the space allotted for dissent is inadequate to fully discuss all of the shortcomings of the Commission’s ID proposal.
HAVA addresses the potential for fraudulent registration by individuals claiming to be someone they are not, and the Report contains no evidence that this reform is not working or that the potential for fraud in voter registration or multiple voting will not be addressed once the States fully implement the HAVA requirement for computerized, statewide registration lists. In fact, it offers scant evidence that this problem is widespread or that such a burdensome reform is required to solve it.
REAL ID is a driver’s license, not a citizenship or a voting card. The Report notes that 12% of the voting age population lack a driver’s license. While it recommends that States provide an alternative photo voting card to non-drivers free of charge, States are likely to require the same documentation that is required of drivers.
The documents required by REAL ID to secure a driver’s license, and consequently a photo ID to vote under this recommendation, include a birth certificate, passport or naturalization papers, a photo identity document, and proof of Social Security number. Obtaining such documents can be difficult, even for those not displaced by the devastation of Hurricane Katrina. For some, the Commission’s ID proposal constitutes nothing short of a modern day poll tax.
Important omissions raise doubts about the completeness of this Report. The lack of a recommendation on counting provisional ballots in Federal and statewide races is unfortunate. Our goal should be to ensure that the maximum number of eligible ballots are counted. Eligibility to vote for President is not dependent upon the precinct in which the voter resides. Similarly, reforms that expand access to the ballot box for working people, the disabled, elderly and minorities, such as early voting and vote-by-mail, are inadequately addressed by this Report.
Election reform must be about empowerment, not disenfranchisement. Raising needless impediments to voting or creating artificial requirements to have one’s vote counted are steps backward. The mere fear of voter fraud should never be used to justify denying eligible citizens their fundamental right to vote.
I am a professor who specializes in election law, and I am writing separately to express my dissenting views to the Carter-Baker Commission’s photo ID proposal. Unfortunately, the Commission rejected my 597-word dissent and allowed me only 250 words (this limitation on dissent was first announced at our final meeting). I believe that the issues before the Commission are of great consequence to our democracy and deserve more discussion. Thus, my concerns with the Commission’s ID proposal and the shortcomings of the Commission’s deliberative process are examined in greater detail at www.carterbakerdissent.com.
Opponents of a voter photo ID argue that requiring one is unnecessary and discriminatory.
Numerous examples of fraud counter the first argument. In 2004, elections in Washington state and Wisconsin were decided by illegal votes. In Washington, this fact was established by a lengthy trial and decision of the court. In Wisconsin, this fact was established by a joint report written by the U.S. Attorney, FBI, Chief of Police and senior local election official – both Republicans and Democrats. In other states, most notably the states of Ohio and New York, voter rolls are filled with fictional voters like Elmer Fudd and Mary Poppins.
Addressing the second concern, the Commission recommendation is for states to adopt safeguards that guarantee all Americans equal opportunity to obtain an ID required for voting. The safeguards include initiatives to locate those voters without IDs and to provide them one without cost. Under the recommendation, eligible voters can cast a provisional ballot that will be counted if they present their photo ID within 48 hours. Far from discriminatory, a mandatory voter ID provides means by which more Americans may obtain the identification already required for daily functions -- such as cashing a check, entering a federal building, or boarding an airplane.
We present this recommendation on a nationwide basis so that states can avoid some of the problems previously highlighted.
I have given the majority of my career to the fair and impartial oversight and conduct of elections, serving 20 years as an elected Secretary of State. It has been an honor to serve on the Carter-Baker Commission and I believe this report is timely, accurate and will provide our country with new ideas to continually reform and improve our elections.
My only exceptions to this report are found in Section 3.1 and Section 4.2. Numerous countries are moving ahead of America in the field of election technology. On voting machines and electronic voting devices, limiting voter verified audit trails only to paper is a mistake. New technology has far greater potential than paper in this arena.
It is my strong belief that the expansion of voting by mail, under strict guidelines to prevent fraud, will ensure that our voting participation will increase dramatically, especially in local and off-year elections.
I support the Commission’s major recommendations, especially those dealing with improved registration systems and the prevention of election fraud. I have reservations about several other proposals, among which the following require specific comment: Recommendations 4.6.1 and 4.6.2. Substantive decisions about criminal penalties are outside the scope of this Commission’s mission, which deals with election administration. Uniformity should not be imposed on the states, some of which may have very sound policy reasons for denying the franchise to all felons or to a larger class of felons than this Commission prefers.
Recommendations 6.1.1, 6.1.2, and 6.1.4. The Commission mistakenly assumes that putatively nonpartisan election administration is necessarily preferable to other approaches. Moreover, the Commission’s proposal to add to the EAC a fifth, putatively nonpartisan member (who would serve as the chair) is profoundly misguided. All the functions that the EAC has, or could sensibly be given, can be carried out under the current bipartisan, four-member structure. If the EAC were reconstituted in the way proposed by this Commission, it would naturally become a magnet for additional functions, and would probably come eventually to serve as a national election administrator, thus displacing the states from their proper role in our decentralized system of governance. I believe this would be a terrible mistake.
Recommendation 7.1.3. This proposal calls for an inappropriate and constitutionally dubious interference with the freedom of the press.
With regard to Recommendation 9.1.1, I agree on the need for regional presidential primaries, but I disagree that Iowa and New Hampshire should come first. At present the barriers to candidates unaffiliated with the major political parties gaining a place on the presidential ballot are substantial. Thus, the primary system is the major way for the American people to participate in the process of selecting candidates for president. But it gives disproportionate influence to those states that go first. One problem with Iowa is that the state decides by a caucus rather than a secret ballot, but the bigger problem with Iowa and New Hampshire is that these states have demographic profiles that make them very different from the rest of the country. Iowa and New Hampshire, according to the 2003 census, have populations that are around 94-95 percent White, while nationally Whites are 76 percent of the population. Hence, the debates are shaped in ways that do not necessarily reflect the interests of minority populations or of our diverse nation.
^ Back to top
FORMER PRESIDENT JIMMY CARTER served as the 39th President of the United States. Among his administration’s accomplishments were the Panama Canal treaties, the Camp David Accords, and the SALT II treaty with the Soviet Union. He began his political career in the Georgia Senate and was elected governor of Georgia in 1970.
In 1982 after leaving the White House, he founded The Carter Center, which he dedicated to resolving conflict, fighting disease, strengthening democracy, and advancing human rights. He received the Nobel Peace Prize in 2002 for his efforts.
FORMER SECRETARY OF STATE JAMES A. BAKER, III has served in senior government positions in three presidential administrations. In 1989, President George H.W. Bush appointed him to serve as the nation’s 61st Secretary of State. During his tenure at the U.S. Department of State, he traveled to 90 foreign countries as the U.S. confronted the challenges and opportunities of the post-Cold War era.
Mr. Baker led presidential campaigns for Presidents Ford, Reagan, and Bush over the course of five consecutive presidential elections from 1976 to 1992. He is presently a senior partner in the law firm of Baker Botts and serves as honorary chairman of the James A. Baker III Institute for Public Policy at Rice University.
ROBERT PASTOR is Director of the Center for Democracy and Election Management, Professor of International Relations, and Vice President of International Affairs at American University. From 1985 until coming to AU in 2002, Dr. Pastor was Fellow and Founding Director of The Carter Center’s Latin American Program and its Election Monitoring Initiatives. He served as President Carter’s representative on the Carter-Ford Commission on Election Reform. He has taught at Harvard University, where he received his Ph.D. in Government, and is the author of 16 books.
BETTY CASTOR was the 2004 Democratic candidate for U.S. Senate in Florida. She has held prominent leadership positions in education, most recently as president and CEO of the National Board for Professional Teaching Standards in Arlington, Virginia. Before joining the National Board, Ms. Castor served as president of the University of South Florida and as Florida Commissioner of Education. She is the founder and president of a political action committee called Campaign for Florida’s Future, dedicated to increasing citizen participation in public life.
TOM DASCHLE served as a U.S. Senator from South Dakota for 18 years and held a number of Democratic leadership positions, including Senate Majority Leader and Senate Minority Leader. Before entering the Senate, Mr. Daschle served four terms in the U.S. House of Representatives and quickly became part of the Democratic leadership. His support for the Help America Vote Act of 2002 helped bring the landmark election reform law to passage in the U.S. Senate. In 2005, Senator Daschle joined the Legislative and Public Policy Group of the law firm Alston & Bird, LLP.
RITA DIMARTINO is the former vice president of congressional relations for AT&T. As AT&T’s in-house resource on Hispanic affairs, she provided guidance to senior management about this growing segment of the population and offered leadership on multicultural issues. In 2002, Ms. DiMartino was appointed principal U.S. delegate to the Inter-American Commission of Women and also principal representative to the Inter-American Children’s Institute. Active at all levels of Republican politics, she was elected executive vice chair of the New York State Republican Committee in 1988.
LEE HAMILTON is president and director of the Woodrow Wilson International Center for Scholars. Prior to becoming director of the Wilson Center in 1999, he represented Indiana’s Ninth District in the U.S. House of Representatives for 34 years. During his tenure, Mr. Hamilton served as chairman and ranking member of the House Committee on Foreign Affairs, chairing the Subcommittee on Europe and the Middle East and the Permanent Select Committee on Intelligence. He is currently serving as co-chair of the National Commission on Terrorist Attacks in the U.S.
KAY COLES JAMES was director of the U.S. Office of Personnel Management from 2001 to 2005. She was a senior fellow and director of The Citizenship Project at the Heritage Foundation, leading efforts to restore a strong ethic of citizenship and civic responsibility nationwide. Ms. James is also the former dean of the school of government at Regent University and has served under President George H. W. Bush as associate director of the White House Office of National Drug Control Policy and as assistant secretary for public affairs at the Department of Health and Human Services.
BENJAMIN LADNER has been President and Professor of Philosophy and Religion at American University since 1994. He chairs the Board of Trustees of the Consortium of Universities of the Washington Metropolitan Area, comprised of fourteen colleges and universities, with 130,000 students. Before coming to American University, Dr. Ladner was president of the National Faculty, a national association of university professors founded by Phi Beta Kappa.
DAVID LEEBRON is President of Rice University in Houston, Texas. He is the former dean of the Columbia University School of Law, where he worked from 1989 to 2004. From 1983 to 1989, he was a professor of law at New York University and director of NYU’s International Legal Studies Program. He is a member of the American Bar Association’s Standards Review Committee and the American Law Deans Association’s Board of Directors. Mr. Leebron has taught and published in areas of corporate finance, international economic law, human rights, privacy, and torts.
NELSON LUND is the Patrick Henry Professor of Constitutional Law and the Second Amendment at George Mason University in Virginia, where he teaches on topics that include federal election law. Mr. Lund served as a law clerk to the Honorable Patrick Higginbotham of the United States Court of Appeals for the Fifth Circuit and to the Honorable Sandra Day O’Connor of the United States Supreme Court. Following his clerkship with Justice O’Connor, Mr. Lund served in the White House as associate counsel to the president from 1989 to 1992.
SHIRLEY MALCOM is head of the Directorate for Education and Human Resources Programs of the American Association for the Advancement of Science (AAAS). The Directorate includes AAAS programs in education, activities for underrepresented groups, and public understanding of science and technology. Dr. Malcom serves on several boards -- including the Howard Heinz Endowment and the H. John Heinz III Center for Science, Economics and the Environment -- and is an honorary trustee of the American Museum of Natural History. She is a trustee of the California Institute of Technology and a regent of Morgan State University.
BOB MICHEL served as a member of the U.S. House of Representatives from Illinois from 1957 to 1993. During that time he held a number of leadership roles, including those of Minority Whip and Minority Leader. Mr. Michel was a delegate to the Republican National Convention from 1964-1992 and permanent chairman of the Republican National Conventions of 1984, 1988, and 1992. He served with the Thirty-Ninth Infantry Regiment as a combat infantryman in England, France, Belgium, and Germany from 1943 to 1946.
SUSAN MOLINARI is the President and CEO of the Washington Group, a government relations and lobbying firm. She was a member of Congress from New York from 1990 to 1997. In 1994, she was elected to the Republican Majority Leadership, making her the highest-ranking woman in Congress. In 1996, she was selected by Robert Dole to be the Keynote Speaker at the Republican National Convention in San Diego, California. Prior to Congress, Molinari was twice elected to the New York City Council, where she was Minority Leader.
ROBERT MOSBACHER is Chairman of Mosbacher Energy Company. He is the past chairman of the Republican National Committee and served as national finance chairman for the election campaigns of Presidents Ford and George H. W. Bush. Mr. Mosbacher served as Secretary of Commerce under President Bush from 1989 to 1992, and was awarded the Aztec Eagle Award from Mexico President Ernesto Zedillo for his role in developing the North American Free Trade Agreement (NAFTA). He is a trustee emeritus for the Aspen Institute for Humanistic Studies and past chairman of the Americas Society/Council on the Americas.
RALPH MUNRO served as Washington Secretary of State from 1980 to 2001. His achievements include implementing a presidential primary allowing independent voters to participate in the nomination process; transitioning election equipment from lever machines to optical scan systems; designing a “Motor Voter” registration system; and supporting a program that allowed Desert Storm troops to vote in Washington elections via fax from the Persian Gulf. Mr. Munro currently serves on the Board of Directors of numerous technology companies, including Dategrity, a provider of secure Internet technology systems.
JACK NELSON is a Pulitzer Prize-winning journalist and former Washington bureau chief for the Los Angeles Times. He covered the past six presidents and every presidential campaign from 1968 through 1996. Since retiring in December 2001, he has taught at the University of Southern California’s School of Journalism. In 2002, Mr. Nelson was a Shorenstein Fellow at Harvard University’s Kennedy School of Government. He was presented the Drew Pearson Award for Investigative Reporting and the Robert F. Kennedy Award for Lifetime Achievement in Journalism.
SPENCER OVERTON is a professor at The George Washington University Law School who specializes in voting rights and campaign finance law. His academic articles on election law have appeared in several leading law journals, and his book "Stealing Democracy: The New Politics of Voter Suppression," will be published and released by W.W. Norton in June 2006. Professor Overton formerly taught at the University of California, Davis, and served as the Charles Hamilton Houston Fellow at Harvard Law School. He currently serves on the boards of Common Cause, the National Voting Rights Institute, and the Center for Responsive Politics.
TOM PHILLIPS is a partner in the law firm of Baker Botts LLP. From 1988 to 2004, he was Chief Justice of the Supreme Court of Texas, and was elected and re-elected four times. During his tenure, he served as president of the National Conference of Chief Justices (1997-98), a member of the Committee on Federal-State Relations of the Judicial Conference of the United States, and an advisor to the Federal Judicial Code Project of the American Law Institute.
SHARON PRIEST is the former Arkansas Secretary of State and the first woman to hold that position. Prior to her election to statewide office in 1994, she has served as mayor of Little Rock. She is also the former president of the National Association of Secretaries of State. Currently, Ms. Priest chairs the Arkansas State Election Improvement Study Commission, the State Board of Election Commissioners, and the Capitol Arts and Grounds Commission. She has also received the TIME/NASBE Award for Outstanding Leadership in Voter Education.
RAUL YZAGUIRRE is presidential professor of practice in community development and civil rights at Arizona State University. He has devoted his career to advocacy issues facing the Hispanic community. He is the founder of Interstate Research Associates, a Mexican-American research association and nonprofit consulting firm. From 1974 to 2004, Mr. Yzaguirre was president of the National Council of La Raza. In addition to his work with La Raza, he helped establish the National Hispanic Leadership Agenda and the New American Alliance, among other organizations.
DANIEL CALINGAERT is the Associate Director of the Center for Democracy and Election Management at American University and Associate Director of the Commission. He has served as Program Director for Asia and Deputy Director for Eastern Europe at the International Republican Institute, where he designed and managed a wide range of programs to promote democracy. Dr. Calingaert previously directed programs to reform social science education at universities across Eastern Europe and Eurasia.
DOUG CHAPIN is Director of electionline.org and Research Director for the Commission. He has worked on election issues for more than 15 years, with extensive experience that includes positions with the Federal Election Commission, the U.S. Senate Rules Committee, and Election Data Services, Inc. Before becoming electionline.org’s first director, he worked at Skadden, Arps, Slate, Meagher & Flom LLP.
KAY STIMSON is the Associate Director of Media and Public Affairs to the Commission. She has served as Director of Communications at both the U.S. Election Assistance Commission and the National Association of Secretaries of State, where she served as the association’s spokesperson and managed its voter outreach efforts. Prior to joining NASS, Ms. Stimson spent more than five years in the field of television journalism as a news anchor and political reporter.
MARGARET MURRAY GORMLY is the Administrative Coordinator for Dr. Robert A. Pastor, Executive Director of the Commission, and as such, she has handled the senior-level administrative affairs of the Commission. She serves as the manager of the Office of International Affairs and has provided senior-level administrative support for the Commission. Before joining the AU staff, she was the executive assistant to the CEO and COO of GW Solutions.
MEEGAN MCVAY is the Grants and Proposals Manager for the Commission. Since 2003, she has served as the primary fundraiser in the Office of International Affairs, working with the Center for North American Studies and the Center for Democracy and Election Management. Ms. McVay is a Certified Fund Raising Executive with more than eight years of development experience, including positions with the Brookings Institution and ACCION International.
PAULINA PUIG is the Web Master for the Commission, managing Web operations for the Office of International Affairs, Center for Democracy and Election Management, and Center for North American Studies. She is also responsible for the Web sites of the AU Abroad and Abroad at AU programs, as well as the ABTI-American University of Nigeria. Ms. Puig previously worked as a technology consultant for a government agency and was a senior Web developer for Discovery.com.
VASSIA GUEORGUIEVA is a Ph.D. candidate in public administration at the AU School of Public Affairs and a Graduate Research Assistant for the Commission. She has worked for the Bulgarian Parliament and the Organization for Security and Cooperation in Europe.
NICOLE BYRD is a M.A. candidate in International Peace and Conflict Resolution/Foreign Policy at the AU School of International Service and a Graduate Research Assistant for the Commission. She is also vice-president of the Graduate Student Council.
JOHN HENDERSON is a Junior Fellow at the Center for Democracy and Election Management and a Graduate Research Assistant for the Commission. As a Rhodes Scholar, he completed a graduate degree in Comparative Politics at the University of Oxford.
KIMBERLY CARUSONE is a M.A. candidate in International Education at American University and the Assistant Web Master for the Commission. She previously worked in publishing and marketing and is a graduate of Pennsylvania State University.
ZACHARY PFISTER is a B.A. candidate in Conflict Studies at DePauw University in Indiana, where he is also the student body president. He was a Summer Research Assistant to the Commission.
The Commission on Federal Election Reform is organized by American University’s Center for Democracy and Election Management.
The Center for Democracy and Election Management, established in September 2002, is dedicated to educating students and professionals about best practices in democracy and conducting public policy-oriented research on the management of elections. In addition, the Center seeks to serve as a venue for public policy discussion on these topics and to provide an institutional base for international scholars to study and teach about democratic processes. Dr. Robert A. Pastor serves as its Director. The Center is part of American University in Washington, DC.
Rice University’s James A. Baker III Institute for Public Policy
The Carter Center
Carnegie Corporation of New York
The Ford Foundation
John S. and James L. Knight Foundation
Electionline.org/The Pew Charitable Trusts
April 18, 2005
American University (Washington, DC)
Panel I: Elections and HAVA: Current Status
Gracia Hillman, Chair, U.S. Election Assistance Commission
Chellie Pingree, President, Common Cause
Kay J. Maxwell, President, League of Women Voters of the U.S.
Henry Brady, Professor of Political Science and Public Policy,University of California
Panel II: Access and Integrity
Barbara Arnwine, Executive Director, Lawyers’ Committee for Civil Rights Under Law
John Fund, Wall Street Journal Editorial Board
Colleen McAndrews, Partner, Bell, McAndrews & Hiltachk, LLP
Arturo Vargas, Executive Director, National Association of Latino Elected and Appointed Officials
Panel III: Voting Technology and Election Administration
Jim Dickson, Vice President for Governmental Affairs, American Association of People with Disabilities
David Dill, Professor of Computer Science, Stanford University
Hon. Ron Thornburgh, Secretary of State, State of Kansas
Richard L. Hasen, Professor of Law, Loyola Law School
June 30, 2005
Rice University (Houston, TX)
Panel I: Voter Registration, Identification, and Participation
Ken Smukler, President, InfoVoter Technologies
Michael Alvarez, Professor of Political Science, California Institute of Technology
Paula Hawthorn, Former Manager of Operating Systems Research, Hewlett-Packard Laboratories
Robert Stein, Dean of Social Sciences and Professor of Political Science, Rice University
Panel II: Voting Technology
Dan Wallach, Associate Professor of Computer Science, Rice University
Beverly Kaufman, Clerk, Harris County, Texas
Special thanks to Harris County, Texas, and the Nevada Secretary of State Office’s Elections Division for providing electronic voting machines that were demonstrated for the Commission during this session.
Panel III: Election Management and Election Reform
Donald J. Simon, Partner, Sonosky, Chambers, Sachse, Endreson, & Perry, LLP
Louis Massicotte, Professor of Political Science, University of Montreal
Norman Ornstein, Resident Scholar, American Enterprise Institute
July 15, 2005
Woodrow Wilson International Center for Scholars
Special thanks to the following Members of Congress for their comments and participation, including related committee staff participation: Rep. Robert Ney (R-OH), Rep. Steny Hoyer (D-MD), Rep. Juanita Millender-McDonald (D-CA), Rep. Rush Holt (D-NJ), and Rep. John Conyers (D-MI).
July 16, 2005
Common Cause Headquarters
Special thanks to Ed Davis and Barbara Burt of Common Cause for organizing this meeting.
August 13, 2005
Beverly Hilton Hotel
Los Angeles, CA
Throughout the course of its research and deliberations, the Commission benefitted greatly from
the substantial contributions of academic advisors and other experts, as well as opinions shared by
citizens around the country. While we wish to acknowledge the distinguished individuals who aided our work, this does not imply that they agree with all of the report’s recommendations.
Nonetheless, their work was invaluable and we want to express our gratitude.
Alan Abramowitz, Professor of Political Science, Emory University
Michael Alvarez, Professor of Political Science, California Institute of Technology
Curtis Gans, Director of the Center for the Study of the American Electorate, American University
Mark Glaze, Director of Public Affairs, The Campaign Legal Center
Paul Gronke, Associate Professor of Political Science, Reed College
Richard Hasen, Professor of Law, Loyola Law School, and Editor of Election Law Journal
Paula Hawthorn, Former Manager of Operating Systems Research, Hewlett Packard Laboratories
Mark F. Hearne II, Partner, Lathrop and Gage LC
Steven Hochman, Research Director, The Carter Center
Robin Leeds, Scholar-in-Residence at the Women & Politics Institute, American University
R. Doug Lewis, Executive Director, The Election Center
David Lublin, Associate Professor of Government, American University
Jennifer McCoy, Director of the Americas Program, The Carter Center
Karen O'Connor, Professor of Government and Director of the Women & Politics Institute, American University
Norman Ornstein, Resident Scholar, American Enterprise Institute
Cameron Quinn, U.S. Elections Advisor, International Foundation for Election Systems
Jamin Raskin, Professor of Law, American University Washington School of Law
Rob Richie, Executive Director, Center for Voting and Democracy
John Samples, Director of the Center for Representative Government, CATO Institute
Leonard Shambon, Counsel, Wilmer Cutler Pickering Hale and Dorr LLP
Richard G. Smolka, Professor Emeritus of Political Science, American University
Robert Stein, Dean of Social Sciences and Professor of Political Science, Rice University
James Thurber, Director of the Center for Congressional and Presidential Studies, American University
Dan Wallach, Associate Professor of Computer Science, Rice University
Tracy Warren, Director, Pollworker Institute
Kimball W. Brace, President, Election Data Services
Craig S. Burkhardt, Chief Counsel for Technology, U.S. Department of Commerce
Stephen E. Fienberg, Professor of Statistics and Social Science, Carnegie Mellon University
Jonathan Frenkel, Director for Law Enforcement Policy, U.S. Department of Homeland Security
John Mark Hansen, Dean of the Social Sciences Division, University of Chicago
Paul Herrnson, Director of the Center for American Politics and Citizenship, University of Maryland
Therese Laanela, Senior Program Officer, International Institute for Democracy and Electoral Assistance
Herbert Lin, Senior Scientist , National Research Council
Thomas Mann, Senior Fellow, Brookings Institution
Robert Montjoy, Professor of Public Administration, University of New Orleans
M. Glenn Newkirk, Principal, InfoSENTRY Services
Jacqueline Peschard, Professor, Universidad Nacional Autonoma de Mexico
John Petty, Chairman, TecSec
Aviel Rubin, Professor of Computer Science, Johns Hopkins University
Robert Saar, Executive Director, DuPage County Election Commission, Illinois
Fritz Scheuren, President, American Statistical Association
Ari Schwartz, Associate Director, Center for Democracy and Technology
Michael D. Siegel, Principal Research Scientist at the Sloan School of Management, Massachusetts Institute of Technology
Hans A. von Spakovsky, Counsel to the Assistant Attorney General for Civil Rights, U.S. Department of Justice
John Thompson, Executive Vice President of the National Opinion Research Center, University of Chicago
Dan Tokaji, Assistant Professor of Law at Moritz College of Law, Ohio State University
Wai L. Tsang, Principal Engineer, TecSec
Tova Wang, Democracy Fellow, The Century Foundation
Hon. Andrew Young, Professor of Policy Studies, Georgia State University
American University staff: Lisa Arakaki, Keith Costas, Maralee Csellar, Clark Gregor, Katherine Kirlin, Todd Sedmak, David Taylor, Leslie Wong, and Julie Weber.
Carter Center staff: Nancy Koningsmark, Faye Perdue, Jane Quillen, and Lisa Wiley.
James A. Baker III Institute for Public Policy staff: B.J. Almond, Charlotte Cheadle, Maggie Cryer, Sonja Dimitrijevich, Kathryn Hamilton, Molly Hipp, and Ryan Kirksey.