As the Supreme Court has observed, the right to vote is the guarantor of all other rights. The laws and policies we enact as a people are heavily influenced by the electoral laws and policies that govern the political process. Sadly, the American electoral system is in many respects an outlier among Western democracies. USAID would not grant unconditional funding to a country whose democratic design looked like ours.

There is fairly widespread agreement among academics and policymakers about some of the fundamental problems with our democracy. We may quibble about which is the best alternative, but most of us agree that politicians should not have the power to draw their own districts, that money has played a corrosive role in the political process, and that the nuts and bolts of our election system are in dire need of repair.

So, working with the Tobin Project, a research organization based in Cambridge, Massachusetts, that connects leading academics and lawmakers, I asked a number of the top democracy scholars in the country for small-scale interventions that would improve the health of our democracy. While academics tend to value ambitious conceptual work and dramatic reform proposals, our goal was to offer members of Congress a diverse array of pragmatic policy proposals—reforms that could be implemented relatively easily without a full-scale retooling of the political system. We met with members of Congress in June in Washington, D.C., to present our group’s proposals. We envision this year’s gathering as part of a one-two punch: now that we have shared our small-scale proposals, in 2007 we will meet again to build on the lessons learned—particularly the feedback provided by members of Congress—and sketch out a blueprint for broader democratic reform.

While the following proposals are informed by a pragmatic approach to politics, don’t let the label “small-scale” fool you. These corrective actions might well have significant long-term and productive effects. They are wedge strategies, efforts to introduce a different dynamic into the political process and to create incentives for better, more productive politics in the future. —Heather Gerken


ONE: Push states to mandate voter registration.

Luis Ricardo Fraga

It is well known that one of the most significant barriers to voting is the requirement that citizens register themselves. In this regard, the National Voter Registration Act of 1993 has achieved considerable progress. States are now required to (1) “provide individuals with the opportunity to register to vote at the same time that they apply for or seek to renew a driver’s license” and “forward the completed application to [the] appropriate . . . local election official,” (2) “offer voter registration opportunities at all offices that provide public assistance,” (3) allow citizens to register to vote by “using mail-in forms developed by each state,” and (4) “maintain a unitary voter registration system for federal and state elections.”

Although some argue that this legislation has virtually eliminated the barrier that registration presents, over one quarter of all eligible citizens are still not registered to vote. According to the Current Population Survey conducted by the Census Bureau in 2004, only 72.1 percent of eligible American citizens reported being registered to vote for the 2004 election. Even more significantly, these rates of registration varied systematically by race and ethnicity. Although 75.1 percent of non-Hispanic white citizens reported being registered, the percentages were 68.7 percent for Blacks, 57.8 percent for Hispanics, and 51.8 percent for Asian-Americans. Stated differently, nearly 55 million American citizens are not registered to vote, and these individuals are disproportionately African-American, Latino, and Asian-American.

Several states have tried to combat this problem. North Dakota has eliminated the requirement that its citizens register in order to vote. Idaho, Maine, Minnesota, New Hampshire, Wisconsin, and Wyoming have enacted a same-day registration system that seems to enhance voter participation. Nonetheless, it is clear that registration is still a major barrier to effective voter participation.

To eliminate this barrier, Congress should pass legislation to provide incentives for states to implement systems of mandated voter registration. Under such a system, a state-level administrative agency, such as the office of voter registration (OVR) in the office of the secretary of state, would be given primary responsibility for registering all eligible citizens to vote. Although in principle this agency could be located within the national government and could provide uniform practices, the mandate should come from state law if we are to maintain consistency with states’ responsibility for voter registration. Participation in the program would be voluntary.

As an incentive for states to participate, the federal government should fund an initial three-year pilot project. The funding could be structured in a number of ways and need not require the federal government to pay the full costs; for example, states’ contributions could be matched two to one. What is important is that the federal government encourage states to participate by providing clear incentives. Each state should also be provided expert technical assistance.

This agency would identify and verify voter eligibility in the following ways and others:

¶ Verify in state databases the address and eligibility of those currently registered to vote; if a person’s address has changed, automatically update the record accordingly.

¶ Ask all persons who formally enroll, receive services, or otherwise solicit assistance from public agencies including schools, hospitals, and social-service agencies if they are citizens of the United States and if they are registered to vote. If a person indicates that she is a citizen but is not registered to vote, the agency will send her name to the OVR for validation and registration.

¶ Following the example of some European countries, including Belgium, conduct home visits. If a person indicates that he is a citizen and not registered to vote, the person’s name will be submitted to the OVR for validation and registration. This is sometimes referred to as a door-to-door enumeration.

¶ All persons who want to register online will be able to do so through a secure Web site maintained by the OVR. Appropriate validation will be conducted.

¶ Any registered voter who wants to revalidate or change his address will be able to do so online. The OVR will then validate the registration.

¶ Every qualified, validated voter will be issued a voter identification card by the OVR. That card will include a photograph and will be secured to inhibit fraud. This card will be reissued every five years as are drivers’ licenses in many states.

If higher levels of voter registration are to be attained, a program jointly initiated by federal and state governments is necessary, especially if significant differences in registration rates according to race, education level, age, and income are to be overcome. Such a federal-state process is fully consistent with common understandings of democratically driven citizen accountability. The U.S. government cannot deny that the current opt-in system leads to the effective disenfranchisement of substantial numbers of American citizens. To maintain such a system is to accept this disenfranchisement.


TWO: Let citizen assemblies draw districts.

Archon Fung

Since 1998 Congressional incumbents seeking re-election have won 99 percent of the time. Partisan gerrymandering is one practice that has made our political system highly uncompetitive. In 2004, 30 states relied exclusively upon members of their sitting legislatures to formulate state and congressional redistricting plans. In almost all of the remaining states, bipartisan commissions were used to determine state or congressional redistricting maps. Both these methods reliably produce incumbent protection plans. When one party dominates the legislature, they electorally entrench that party. When neither party dominates, or when bipartisan commissions are used, redistricting plans often protect incumbents of both parties by dividing the electoral spoils.

Many reformers favor independent commissions—such as those used in Arizona—as a remedy for this gerrymandering. However, it may be difficult to inoculate such commissions from partisan influence. Most of the notables who would be appointed to such blue-ribbon affairs would likely have political histories and established loyalties. Those who did not would likely be the targets of intense partisan pressure and subterfuge. Even if commissions were politically immunized, they could still be summoned to serve highly partisan ends, as in the recent attempt to initiate redistricting by California’s Republican governor. Furthermore, electoral districting is never merely a technical exercise. Ethical choices must be made. For example, is the coherence of communities more important than competitiveness and political accountability? Appointed experts have dubious democratic standing when it comes to such decisions.

Consider an alternative method of redistricting in which ordinary citizens formulate redistricting plans. The main benefits of this directly democratic alternative are:

Fairness. The formulation of electoral boundaries would be insulated from partisan and incumbent influence.

Ethical transparency. The democratic values and rationales for electoral boundaries would be transparent, explicit, and determinative.

Democratic legitimacy. Electoral boundaries created by ordinary citizens rather than political elites or independent “experts” would build democratic legitimacy for, and popular ownership of, a political system that is now regarded with justified cynicism.

The idea of citizen assemblies as an element of electoral reform is not new. When the western Canadian province of British Columbia recently considered updating its electoral system—the question there was not redistricting but proportional representation—the Liberal Party created an unprecedented Citizens’ Assembly rather than convening a constitutional convention or delegating the task to a legislative committee. The Citizens’ Assembly was composed of 160 citizens who were randomly selected from provincial voting lists. In order to assure a degree of descriptive representativeness, selection was stratified by region and gender. Each member dedicated approximately 30 days to the assembly process over the course of a year.

In their deliberations, members decided that British Columbia’s electoral system ought to serve three fundamental values: fairness, understood as proportionality in the allocation of legislative seats; local representation, understood as the bond between an elected representative and her geographic constituency; and choice, understood as the number of candidates and parties from which voters select. At the end of the process, assembly members voted between two alternatives: a mixed-member proportional (MMP) system and a version of the single transferrable vote. The single transferable vote defeated the mixed-member proportional system by an assembly vote of 123 to 31. Bypassing the legislature, the citizens of British Columbia considered this recommendation in a provincial referendum in May 2005. The measure obtained a majority in all but two of 79 constituencies, but its overall majority of 57.9 percent fell just short of the required 60 percent super-majority threshold. The Liberal Party government has scheduled a second referendum on the single transferable vote proposal for the 2008 election with the hope that interim public discussion will generate a more conclusive decision.

The Citizens’ Assembly has already drawn substantial attention from other provinces in Canada and countries where reformers have been attracted by its direct democracy and populism. Ontario, for example, has reproduced the British Columbia Citizens’ Assembly in almost all of its details. In the Netherlands, a citizens’ assembly of 140 members is convening to consider its electoral system. Legislators in California and the United Kingdom have introduced bills that would introduce the citizens’ assembly concept.

In the United States, states should experiment with citizens’ assemblies as an alternative to legislative committees and independent commissions in both state and congressional redistricting decisions. For congressional redistricting, these experiments might be urged through a federal statute. The muscle to press for state-level redistricting could come from reform-minded legislators and clean-elections organizations in states such as Arizona, Vermont, Maine, Massachusetts, and California.

State citizens’ assemblies should be large enough to ensure some kind of descriptive representation but small enough to make deliberation manageable: they should consist of perhaps 50 to 150 citizens. They might follow British Columbia’s model in selecting an equal number of citizens from each electoral district. Participants should be randomly selected from appropriately stratified lists of eligible voters. And they should be charged with two main questions: first, what are the democratic values that an electoral map should advance and protect (and what is their relative importance)? Such values might include political competition, community representation, minority representation, compactness, political equality, and legislative stability. Second, which electoral map best advances those interests?

State-level citizens’ assemblies should meet frequently—perhaps each weekend—for several months. Their deliberations should be broken into three stages. First, participants should learn the basics of elections and political districting. These include the status quo arrangements; the various outcomes at stake, such as competitiveness and political equality; constitutional and statutory requirements; and the history of districting in their particular state. In the second phase, participants should try to reach a consensus about the values that should guide districting. Third, participants should agree upon a map of electoral districts. They will need the help of professional staff and computer software to produce a good and workable map. Finally, each assembly, assisted by an experienced staff, should issue a report that describes their deliberations and findings.

If an assembly reaches consensus, their findings should be moderately empowered. Their recommendations should either be voted upon by the state legislature without amendment or ratified by popular referendum.


THREE: Create a national voter-registration list.

Samuel Issacharoff

The heated events in Florida in 2000 and Ohio in 2004 brought to light once again the problem of facilitating voting by those eligible to vote. One cause of this problem is the well-known and understandable difficulty of maintaining accurate voter-registration rolls: the American population is highly mobile, with significant numbers moving regularly across electoral districts and oftentimes across state lines. Thus, voter lists regularly become obsolete, voters register simultaneously in multiple states, and the election-day administrative burden becomes increasingly difficult to bear.

Absent federal intervention, efforts among the states to coordinate with each other are often frustrated. At present, a citizen who moves from one state to another may re-register within a short period of arriving, but there is no formal mechanism to ensure that her name is withdrawn from the rolls in her previous state of residence. Similarly, a citizen who lives in Ohio but winters in Florida may very well find himself on the rolls in both places should he decide to opt for Florida as his primary residence and fail to notify Ohio of that fact. The effect is to make it difficult for states to maintain accurate lists, a failing that may result in contested eligibility claims on election day.

The good news is that the Constitution provides a mechanism to help states coordinate a system of better election practices by using congressional elections as a source of federal authority. The specific provision is found in Article I, Section 4, of the Constitution:

The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of choosing Senators.

Under this provision, the authority of Congress is plenary in settivng the manner of holding congressional elections, something that is reinforced by the Supreme Court’s holding in U.S. Term Limits v. Thornton. Federal authority over congressional elections could be used as a way to compel states to use more responsive processes in both state and federal elections.

The proposal here is to federalize the voter-registration lists for federal elections. A federal list that is updated for each congressional election cycle could serve as a template for updating state lists with the addition of new registrants and the removal of those no longer eligible to vote. Only the federal government has the authority to compel (or induce) state compliance with a centralized, regularly updated voter-registration list. As with the requirement of a federal census, the expectation is that federalized registration lists would erode any state interest in independent voter-registration practices and allow a simple and uniform system of identifying eligible voters. Maintaining two voter-registration systems, much like maintaining two census systems, is wasteful and provokes uncertainty when disparities arise. The states came to realize that having one national, federally funded census with binding results would be in everyone’s best interests. The expectation is that the transition to a centralized voter-registration process would be similar to that of the census, with the federal government acting as organizer and custodian of the master list.

It is not difficult to imagine a federal system for federal elections. Each state would be required to submit electronically to a federal election authority, such as the Federal Election Commission, which would create a national election database. The list would be closed for a fixed period before each federal election, for example 15 days before each primary or congressional election. All voter registrations would have to be reported electronically to the federal authorities, with a specific designation for first-time registrants and changes of address. A change of address would trigger removal of the earlier address, a mechanism that would be particularly useful if the voter were moving across state lines.

Once the national list was communicated back to the states, there would be a one-week period for individuals and political parties to check the lists for omissions. There would be a one week period in which to make corrections through state voter-registration officials. After that week, the list would be closed. All citizens on the list would be eligible to vote, and all citizens not on the list would be ineligible. No election-day challenges to voter eligibility as determined by the list would be permitted. While some voters might still mistakenly report to the wrong polling place, the voter rolls would be settled before election day, and eligibility challenges at the polls would be eliminated.

Two other features should be noted. First, the federal statute could criminalize attempts to vote by persons impersonating a registered voter. Second, the federal statute could criminalize attempts to impede a registered voter from voting. These measures would help deter voter fraud and voter harassment.

In addition, the federal statute could define the requirements for voter identification. These might take the form of any government-issued identification card or passport, or utility bills in the name of the voter. If a standard identity card is required, the underlying statute could provide the mechanism for the issuance of such cards at government expense to all citizens.


FOUR: Find more opportunities to introduce underrepresented groups to voting.

Pamela S. Karlan

A bedrock principle of due process is that the kind of notice the government gives someone before depriving her of life, liberty, or property should be the sort that “one desirous of actually informing” the individual “might reasonably adopt.” A “mere gesture” is not enough.

What would happen if we applied this principle to voting? When the government is truly invested in whether a citizen fulfills an obligation—whether registering for the draft, staying clean on parole, or showing up for jury duty—it acts quite differently than it does with respect to political participation. By taking the same active approach to voting, we might increase turnout, particularly among underrepresented groups.

Young people, for example, tend to register and vote at lower rates than their older cohorts. This trend can skew governmental policy in various ways (compare, for example, Congress’s responsiveness to the health-care needs of the elderly with its failure to adapt student-loan programs to respond to rising tuition costs).

There are a variety of circumstances under which young citizens at or approaching voting age encounter the government. Under the National Voter Registration Act (the so-called Motor Voter law), one canonical event—appearance at a state department of motor vehicles office—now triggers a requirement that the state also provide opportunities for voter registration

Here are five suggestions for other events that ought to provide registration opportunities:

Registration for the draft. Under the Selective Service Act, virtually all young men in the United States are required to register for the draft when they turn 18 and are required to notify the government within ten days of any change of address until the year in which they turn 26. This registration requirement is both too inclusive, since it applies to noncitizens and some individuals who are otherwise ineligible to vote, and not inclusive enough, since it excludes women. Nonetheless, the symbolism of linking voter registration with registration for the draft makes this an appealing candidate, and one entirely within federal control. The forms used for Selective Service registration and change of address could be amended to make them parallel the check-off system now used in DMVs in which individuals are asked whether they wish to register to vote as part of the process of seeking a driver’s license and in which changes of address automatically trigger transfer of voter registration.

The student-loan process. Large numbers of students pursuing higher education rely on federally subsidized or provided student loans. As part of the loan process, students have to fill out applications, and as they leave school they must participate in exit interviews and provide information related to loan payment. Again, as with registration for the draft, the category of individuals taking out student loans does not overlap precisely with the category of individuals eligible to vote. But here, too, the government has a natural opportunity to make voter registration easy and transparent.

High-school graduation. The No Child Left Behind Act and related changes in public education now mandate a series of high-stakes tests in which students can receive high-school diplomas only once they have demonstrated mastery of basic subjects. Perhaps the government should adopt a No Voter Left Behind codicil in which high schools are obligated not only to test their students—in many states on civics and American history and government—but also to help them register. Moreover, since schools are often used as polling places, students could get hands-on experience with democracy—perhaps serving as poll workers (a group whose average age is commonly thought to be somewhere way past retirement), and certainly by learning how to use the voting technology employed by their jurisdiction.

Discharge from juvenile supervision. As far as I know, no state disenfranchises individuals solely for contact with the juvenile justice system (unless they are tried as adults and then sentenced to youth facilities). By helping these young people to register to vote when they are discharged, we would be sending a powerful message about their opportunity for a fresh start.

Offender re-enfranchisement. Disenfranchising large numbers of individuals upon conviction of a crime is a distinctively American practice. And the state laws governing the restoration of the right to vote are anything but uniform. Many individuals—ex-offenders, public officials such as parole and probation officers, and local officials responsible for voter registration and conducting elections—are ignorant of the laws in their particular jurisdiction. Some state laws require individuals convicted of federal crimes to have their civil rights restored through federal action—essentially, a presidential pardon—before they can vote once again.

I propose, first, that state and federal corrections authorities be required to inform offenders, upon release from custody or supervision, what their jurisdiction requires for re-enfranchisement. In jurisdictions where eligibility to vote is automatically restored upon release from incarceration or the end of probation or parole, officials should be required to provide registration materials to the offenders. Second, Congress should enact legislation providing that states that receive federal funds for either their prison system or their election system—in other words, every state—cannot condition an individual’s right to vote on his or her receipt of a presidential pardon. Instead, they must make restoration of voting rights to individuals convicted of federal crimes available on the same terms as they do to individuals convicted under their own criminal codes. So in states that automatically restore civil rights for individuals convicted in their courts, restoration would be automatic for federal offenders as well; in states that have a pardon or restoration process, that process could be pursued by federal offenders as well.

FIVE: Reaffirm in law that voting is a fundamental right.

Alex Keyssar

The election-reform community has, in recent months and years, expressed considerable concern about the passage in several states of new, more stringent requirements that citizens present identification documents in order to vote. As a response to that much-warranted concern and to a broader need to protect the right to vote of American citizens, I would propose that Congress pass the following legislation:

The Congress of the United States affirms that the right to vote in elections is a fundamental right of all citizens. If, in order to exercise that right, the citizens of any state are required to present or possess documents of identification, then it shall be the responsibility of that state to ensure that all of its voting-age citizens possess such documents. The costs for the preparation of such documents shall be borne by the state.

A bill of this type would serve several purposes. It would affirm that voting is a right and identify that right as fundamental. To be sure, the notion that Americans possess a right to vote ought to be enshrined in the Constitution, and efforts to make that happen deserve our energetic support. But a constitutional amendment (such as the one sponsored by Congressman Jesse Jackson Jr.) is not going to be passed any time soon. Meanwhile, statutory affirmations should possess some rhetorical (and political) value, and they may have far-reaching legal implications as well.

Indeed, this proposal is designed to illustrate one of the ways that the formal acknowledgement of voting as a “right” could foster concrete applications. The logic of the proposal is that if a state imposes procedural constraints on the exercise of a right, it is the obligation of the state to ensure that those constraints do not prevent citizens from actually exercising that right. If the state creates the constraint, then the state must enable eligible citizens to meet the procedural requirement. That in itself would seem to be a principle worth defending.

This proposal is also designed to shift, at least partially, the terms of the debate about ID requirements—a debate that, in recent months, has focused on Georgia’s new law obliging voters to possess and present government-issued photo-identification documents. Many members of the reform community (myself included) see such requirements as an effort to hinder the electoral participation of some voters, particularly among the poor, the elderly, and the young (in general, those who are least likely to possess drivers’ licenses). But to argue against ID requirements per se is to invite the counterargument that such documents are needed in order to prevent fraud. The debate about the desirability of these requirements then hinges on the record of past fraud and the likelihood of future fraud; for instance, some have argued that Georgia does not need an ID requirement because there is no recent track record of electoral fraud in Georgia.

True enough, but those are not the grounds on which we should be fighting this particular battle: it would be foolish to argue that fraud will never occur, since sooner or later instances of major fraud will occur somewhere. When that happens, the argument that voters need to produce identification documents will be significantly strengthened, while opponents of such laws will appear naive.

Thus, rather than debate the need for ID requirements, it would be more fruitful to focus discussion on a different issue: determining whose responsibility it is to ensure that all citizens are able to meet any state-imposed procedural requirements for casting their ballots. To place that responsibility squarely on the shoulders of the state would encourage (if not compel) state, county, and local officials to be assertive and systematic in their implementation of ID requirements. It would also offer citizens some means of redress—at the least, perhaps, the ability to cast provisional ballots—if they were prevented from voting because they lacked documents. If a voter arrived at the polls without the requisite identification, that would be treated not as the voter’s error (presumably punished by not allowing him to vote), but rather as a failure of the state’s administration of its own law. This approach would be consistent with the existing approach to provisional ballots embedded in the Help America Vote Act.

The compliance of states with a law of this type could be monitored and measured by the Election Assistance Commission, the federal agency charged with administering the Help America Vote Act. (That the EAC is already partially engaged with such issues is made clear in its advisory, dated September 13, 2005, entitled “Provisional Voting and Identification Requirements.”) Sanctions for violating this law could (and perhaps should) be spelled out by Congress; Congress could also choose to appropriate funds to help states conform to this law. Doing so would likely encourage states to adopt photo-ID requirements while moving the country further down the path toward a system of national photo IDs.


SIX: Help the non-wealthy to participate in funding campaigns.

Spencer Overton

Even in the aftermath of recent campaign-finance reforms, a small, wealthy, and homogenous donor class continues to make relatively large contributions that fund the bulk of American politics. Less than two percent of the U.S. population make financial contributions of over $200 to federal candidates. These larger contributions represent the vast majority of funds that candidates receive from individuals, and effectively determine which candidates can afford to run viable campaigns.

Those who make contributions over $200 are also unrepresentative of our nation’s population. While 87 percent of the general public belong to households with annual incomes under $100,000, 14 percent of such donors do. While 76 percent of the general public aged 25 and older lack a college degree, 16 percent of such donors do. And while 31 percent of the general public are people of color, four percent of such donors are.

Widespread participation should be the primary goal of campaign-finance reform. Instead of “preventing corruption” or “equalizing resources among candidates,” future campaign reforms should reduce the impact of disparities in wealth on political participation and empower more citizens to participate in the funding of campaigns. Specifically, reforms should work to increase the number of people who give contributions of $100 or less to candidates, parties, and PACs; and to increase the importance of small donors so that contributions of $100 or less account for at least 75 percent of the money raised by most candidates, parties, and PACs.

Lawmakers should provide incentives for a broader and more diverse group of Americans to make small contributions and enhance the effect of their participation by adopting matching funds and tax credits for contributions of $100 or less. Congress should adopt such a system for contributions to all federal parties, PACs, and presidential and congressional candidates when it considers revising the federal presidential funding system. The cost would be $1.5 billion annually, or a one-time allocation of $63 billion to establish a perpetual trust.

One reform would provide a four-to-one match for contributions of $100 or less. In other words, if a contributor gives $100 to a candidate, the candidate would receive another $400 in public funds, producing a total contribution worth $500 to the candidate. This reform would replace the existing presidential public-financing system and would improve on the current system by limiting matching funds to smaller contributions, encouraging citizen engagement through the general election, and expanding public financing to congressional elections. All federal candidates listed on the ballot would be eligible to receive matching funds that correspond to their support by contributors of $100 or less, since the match is designed to help citizens support their candidate of choice rather than equalize funds between candidates.

A second reform would provide a tax credit for smaller contributions. For example, a contributor who gives $100 to a candidate would receive a tax credit that reduces her tax bill by $100. The tax credit would be limited to $100 per year (or $200 for joint returns), refundable so that poorer families who do not pay taxes could use it, and well-publicized to ensure that its use would not be limited to those who can afford attentive accountants.