An Agenda for Election Reform

As the dispute over the 2000 presidential election results in Florida extended through November and into December, many Americans were horrified to learn about the seemingly chaotic and flawed manner in which federal elections are administered. Calls to upgrade the nuts and bolts of the election system attracted broad support, but policymakers confronted formidable substantive, legal, and political hurdles that slowed federal action to a crawl.

Now that the states are moving to alter their systems and there is growing consensus on reform, Congress should move to define and finance the federal role. This brief identifies a strategy and set of priorities for the federal government to assist in a national effort to improve the election system.


The extraordinarily lengthy and controversial resolution of the 2000 U.S. presidential election provided tens of millions of Americans with a national civics lesson. For some, it was an introduction to the mechanics of the Electoral College, and the reality that the national popular vote does not always square with the only count that matters—the electoral vote. For others, the dead-heat finish, in which several hundred ballots in Florida determined the outcome of an election in which over a hundred million ballots were cast, led to sober reflection on the responsibilities and potential power of individual citizens. But for most, the striking and troubling feature of the lesson was the discovery of the underside of American democracy—the highly decentralized, non-uniform, antiquated, confusing, error-prone, under-budgeted, poorly-staffed, arbitrary, and politicized manner that is too much a part of the way that federal elections are administered. Reports from Florida indicated that tens of thousands of eligible voters, many times the margin of victory of George W. Bush, were disenfranchised because their names were improperly purged from registration rolls or because their ballots were spoiled. What Americans had taken for granted—that those who have a right to vote are able to do so, and have their votes counted accurately—was now seriously open to question.

The Florida election debacle, and the dawning realization that serious problems casting and counting votes exist throughout the country, produced a clarion call to document the flaws and take decisive steps to ensure that future elections are administered consistent with the very highest standards Americans expect of their democracy. Major news organizations produced a series of informative reports on the nuts and bolts of elections that had seldom before captured the attention and resources of editors and producers. At least eight major lawsuits were filed to address the problems experienced by different groups of voters in November. A flood of legislation—calling for studies, public subsidies, and new standards for voting equipment, registration rolls, and vote counts—was introduced in Congress and in the 50 state legislatures. Dozens of national and state commissions, study groups, and committees were charged with analyzing shortcomings in the election system and formulating recommendations for improvement. The stage appeared to be set for a major public policy initiative to deal seriously with the widespread flaws revealed by the 2000 election.

Stalled Reform Efforts

Not surprisingly, the pace of election reform slowed and prospects for speedy enactment of major legislation dimmed as policymakers confronted a host of substantive, legal, and political complications. Early talk of a uniform national ballot quieted as the constraints of the federal system became apparent. Hopes for a quick technological fix faded as contradictory evidence on the impact of different types of voting equipment across the country was uncovered and as critical human factors were identified. Proposals to increase turnout by easing the burdens of voting were seen to entail costly tradeoffs with the timeliness and accuracy of the vote count and the security and privacy of the ballot. The legal obstacles proved no less daunting. Authority over the administration of federal elections is widely dispersed. As eager as federal officials may be to improve the election system, state and local authorities are just as determined to retain the freedom to adopt and administer election systems that respond to local conditions and the preferences of their constituents. Moreover, uncertainties about the precedent of the highly unusual equal protection arguments of Bush v. Gore led to confusion and hesitancy among policymakers.

Most visible are the political hurdles. In Rashomon-like fashion, the major political parties see two starkly different worlds in viewing the shortcomings of the election system. Democrats tend to worry about citizens being denied the right to vote, while Republicans voice concern about fraudulent votes cast by those ineligible to do so. Both parties are clearly influenced by calculations of political self-interest. As with campaign finance, each fears that changing the rules of election administration might systematically advantage the other.

At the same time, President Bush showed little interest in leading a major federal initiative to upgrade the election system. His legislative priorities—tax cuts, education reform, energy development, Social Security restructuring, and national missile defense—flowed from promises made during his campaign, a time when election reform was not on his or anyone else’s radar screen. Competition from these issues for agenda and budgetary priority rendered election reform an afterthought at best. At worst, Bush’s obvious desire to have the country move past the disputed Florida vote count provided him ample incentive to try to keep election reform off the agenda.

So it should come as no surprise that the 107th Congress has dragged its feet on election reform in its first months. But it would be a mistake to conclude that the window of opportunity has closed and the issue is dead. Several states—including Maryland and Georgia—have moved quickly to set new standards for voting equipment and election administration, and in some cases, to provide financial support for machines, poll worker training, and voter education. In early May, the Florida legislature approved a major overhaul of their election system, including a prohibition on punch-card ballots; precinct-based voting technology; subsidies for equipment modernization, a statewide centralized voter registration database, voter education, and training for poll-workers; a standardized ballot design; provisional ballots; and uniform standards for manual recounts. Dozens of others states are on the verge of enacting significant changes in election law and administration.

Viewed from this broader federal perspective, signs of progress are encouraging. Heightened public awareness, the mobilization of advocacy groups, the leverage of lawsuits, and the emergence of new knowledge about best practices has put pressure on state officials to act. While the 5-4 Supreme Court majority in Bush v. Gore tried valiantly to limit the reach of their decision to the specific circumstances of the 2000 presidential election, litigants will aggressively argue to lower courts that disparities within states on how votes are cast and counted violate equal protection rights of disadvantaged citizens. Blue-ribbon commission reports unveiled this summer and fall will generate renewed media attention, as will completion of the National Opinion Research Council coding of all undervotes and overvotes in the Florida election.

Whether and when these and other developments are sufficient to overcome the many obstacles discussed above and generate serious deliberative action in Congress is uncertain. The policy process, especially on issues of political reform, is notoriously complex, conflictual, and time-consuming. When that day arrives, it is crucial that Congress have a clear sense of the comparative advantage of the federal government in strengthening the election system and how it might most constructively complement efforts already underway across the country.

How Not to Proceed

Since Congress has broad authority under the Constitution to regulate the manner of House and Senate elections, to protect the right of citizens to vote, and to initiate amendments to the Constitution altering the method by which presidents are elected, its agenda for election reform can be as ambitious as it desires. But prudence dictates that Congress demur on two major lines of reform. The first is replacing the Electoral College with a direct popular vote. While this is an important subject, worthy of an informed and lively national debate, it ought to be discussed on a separate track. The course of constitutional amendment is long and perilous, and any proposal to eliminate the Electoral College will face insurmountable opposition for the foreseeable future.

The second temptation that should be resisted is to attempt to create a federal election system to administer federal elections, one that includes uniform national ballots, voting equipment, and counting procedures. In this case, the limits and possibilities of the U.S. federal system dictate a less ambitious course. State law now sets many of the terms under which elections are conducted, including how candidates gain access to the ballot, which voting equipment is certified for use by local jurisdictions, whether absentee balloting, voting-by-mail, and early voting are permitted, if straight-ticket party voting is facilitated or discouraged, and how recounts are conducted. States presently differ widely on which minor-party or independent presidential candidates are listed on the ballot, what other state and local elections (including initiatives and referendums) are scheduled at the same time as federal elections, and the methods and times at which ballots may be cast and counted. The cost of federal preemption of this state authority, even if it were possible, could easily outweigh the benefits. Even in the realm of voting equipment, there is much to be said for state and local experimentation with new technologies, which can make a virtue of non-uniformity and facilitate the development, evaluation, and diffusion of innovations.

A Constructive Federal Response

Congress should consider two new or enhanced roles for the federal government on the mechanics of voting. First, the responsibilities and resources of the Office of Election Administration within the Federal Election Commission should be greatly expanded and moved into a new, nonpartisan, independent election administration agency. In addition to establishing technical standards for voting equipment, this agency should sponsor research and development on voting technology, gather systematic data on the performance of various election systems, investigate best practices on all aspects of election administration, develop voluntary standards based on those practices, and serve as a clearinghouse for research on the casting and counting of votes in federal elections.

Second, a federal grant program, administered by the new election administration commission, should be established to help state and local governments upgrade all aspects of their election systems. Given America’s stake in a well-functioning and legitimate election system, and the inevitable under-investment in registration list maintenance, voting machines, poll workers, and voter education by local and state governments, the federal government has the responsibility and ample incentive to provide the financial resources sufficient to ensure that these state systems are brought up to standard. An allocation of $500 million per year is a reasonable place to begin.

First Things First: Voter Registration Lists

The federal government’s highest priority should be encouraging and assisting states that have not yet done so to develop statewide electronic voter registration databases, linked to local registration offices, polling locations, and public agencies involved in the implementation of the National Voter Registration Act (e.g. departments of motor vehicles, public assistance offices, and the U.S. postal service).

Registration rolls are a mess in many states and counties, with error rates as high as 25 percent or more. Listings for registrants who move out of the jurisdiction, die, or become ineligible for some other reason are often not removed in a timely fashion. This increases the risk of fraud. On the other hand, as we saw in Florida in 2000, efforts to purge the rolls of ineligible voters can lead to the erroneous removal of qualified citizens. Long lines and confusion at polling places and disenfranchised voters are the result.

Information-age technology, especially when combined with a unique voter identification number, can make a huge difference in facilitating new voter registrations, keeping registration rolls accurate, reducing election fraud, assisting election officials at polling places, and helping eligible citizens avoid being turned away from the polls on election day. Computerized state voter registration systems also have the virtue of making it easier to move the registration deadline closer to election day.

Voting Equipment

The federal government should also help states and counties upgrade their voting equipment by providing technical assistance and financial support. The oldest forms of voting since the introduction of the Australian ballot in the 1880s—paper ballots and lever machines—have been declining in number and now are used by less than a fifth of the electorate. They were replaced initially with punch cards but increasingly by optical scan and direct recording electronic (DRE) equipment. The 2000 election controversy surrounding the confusing “butterfly ballot” in Palm Beach County and the evidence of a disproportionate number of unmarked, spoiled, or uncounted ballots in Florida precincts with punch card (Votomatic) systems has fueled a growing consensus that punch card voting technology is obsolete and should be replaced as rapidly as possible.

Reality is a bit more complicated. The level of error (intended votes not properly recorded or counted) varies with the design of the ballot, the clarity of instructions, familiarity with the voting equipment and ballot form, the availability of feedback to voters about possible errors in their ballots, and policies governing whether election officials clean and correct obvious errors on ballots before they are counted. Some steps can be taken to minimize the shortcomings of existing equipment. Simply replacing equipment with newer technologies offers no guarantee of error-free casting and counting of ballots.

That is not to argue against replacing punch-card voting systems (or, for that matter, paper ballots and lever machines). On balance, it makes sense to move toward the newer technologies of optical scan and DREs. But voting equipment must be selected with an eye toward the special needs of localities, such as multiple languages and lengthy ballots. Precinct scanning of optical scan forms and ballot review for electronic machines, which allow voters to see if their ballots are properly marked, should be considered an essential feature when acquiring new voting systems. And serious attention must be paid to designing ballots, providing access to disabled citizens, and developing procedures that help citizens properly utilize the voting equipment.

Voting Procedures

These recommendations for upgrading registration lists and voting equipment are based on the premise of election day voting in thousands of polling places across the country. Yet increasingly, this model of a community-based election day experience is being replaced by remote voting over an extended period of time. Oregon now conducts its elections entirely by mail. The state of Washington, with its liberal policy on absentee voting, may not be far behind. Many states have seen a dramatic growth in the percentage of absentee ballots cast, as liberalized laws are quickly exploited by parties and candidates to turn out their supporters in advance of election day. Early voting at polling places is also now permitted in a number of states and localities.

Motivations behind these changes in voting procedures vary, but all share the assumption that making voting more convenient will increase turnout. The evidence to date does not support that assumption. It appears that citizens who avail themselves of these more convenient forms of voting are the ones most likely to show up at polling places on election day. There are important exceptions. Some absentee voters are physically unable to cast ballots in their precincts on election day because of age, physical impairment, or travel commitments, or because they are stationed overseas. But these are the traditional absentee voters, who comprise a shrinking share of all ballots cast before election day.

There are costs associated with these newer forms of mail and early voting: the loss of a shared community experience; the irrelevance of late-breaking campaign events; the threat to privacy provided by secret ballots at polling places; the risks of vote buying and selling; and complications for the timely counting of ballots. Remote Internet voting from home or work incurs these costs and many additional ones. As summarized by the recent Internet Policy Institute report of a workshop sponsored by the National Science Foundation, “Remote Internet voting systems pose significant risk to the integrity of the voting process, and should not be fielded for use in public elections until substantial technical and social science issues are resolved.”

States would be well advised to slow this rush toward new voting procedures, and instead invest in making election day precinct voting more accessible and user friendly. That means more polling places, more convenient locations, additional and better-trained poll workers, electronic links to central voter registration files, clearer instructions to voters, and longer hours for those whose work schedule permits only very early or late voting. The latter recommendation, if coordinated across the country, could provide a uniform poll closing time (e.g. 6:00 a.m. to midnight on the east coast, 6:00 a.m. to 9:00 p.m. on the west coast), thereby constraining the television networks’ practice of projecting the winner of some states while others are still voting.

Voter Education

No aspect of the electoral process in the United States suffers from greater under-investment than preparing citizens to cast their ballot. Mistakes made by voters account for a substantial portion of spoiled ballots and other errors on election day.

Every level of government—national, state and local—has a clear interest in developing and financing programs that educate citizens about their rights and responsibilities as voters. Registration instructions and confirmations, polling place locations and hours of operation, sample ballots, voter guides, and voting equipment demonstrations should all be provided to citizens on a timely basis. Civic, educational, political, and media organizations should do more to educate citizens on the nuts and bolts of voting. Poll workers should be trained to assist voters, provide provisional ballots when registration status cannot be confirmed, and permit voters who spoil their ballot to receive a replacement. By increasing knowledge about when, where, and how to vote, these educational efforts might have the fortunate side effect of engaging more citizens in the democratic process.

Uniformity and Equity

One of the most controversial aspects of the disputed Florida vote count was the alleged disparate impact on racial, age, and partisan groups of the mechanics of registration and voting. Wide variation across the state in voting equipment, ballot design, accuracy of registration lists, number and training of poll workers, lines at polling places, the availability of provisional ballots, the treatment of absentees, and procedures for counting and recounting ballots appear to have advantaged some groups at the expense of others. In the case of the infamous Palm Beach County butterfly ballot designed, ironically, by a well-intentioned Democratic election official, the statistical evidence is overwhelming that many thousands more Democrats than Republicans were confused and spoiled their votes for president.

Florida has now adopted a new election system that will eliminate much if not all of this variation among its local jurisdictions. One can argue that the equal protection argument of Bush v. Gore, although intended by the Supreme Court majority to be limited to manual recount standards in Florida in the 2000 presidential election, provides a powerful leg in response to special needs of localities may be a strength, not a weakness, of the election system. But states will have to grapple with this issue, and make serious efforts to impose certain uniform standards that constrain the discretion of local election precedent for demanding similar actions by the other states. How much uniformity within states is necessary to avoid potentially harmful disparate impact on groups of voters is not at all clear. Some variation

For its part, the federal government can assist this effort by developing a reliable body of knowledge about the performance of voting equipment, the impact of election laws, and the efficacy of alternative approaches to the administration of elections. It can also pay its fair share of the cost of administering federal elections, and create a new independent agency in Washington to do the continuing work necessary to ensure that we will never again have to discover and explain to ourselves and the rest of the world, as we did in 2000, the underside of American democracy.