From:John J.A. Burke and Judith Ungar

Memorandum

To:The Commission

From:John J.A. Burke and Judith Ungar

Date:December 3, 2001

Re:Election Law

The Miller Center of Public Affairs published a Task Force Report entitled To Assure Pride and Confidence in the Electoral Process (August 2001) (hereafter the Report) to accompany the National Commission on Election Reform.[1] The Report was prepared under the direction of John Mark Hansen of Harvard University, was compiled from data collected from state election officials, culled and summarized by staff headed by Michael A. Neblo, Robert Wood Johnson Fellow in Health at the University of Michigan, and contributions from analysts, law practitioners and various scholars, including professors at the University of Chicago, University of California at Berkeley and the Massachusetts Institute of Technology.

The Report provides information about current practices in federal elections based upon the laws of every State. It then analyzes those practices and makes recommendations for reform. The Report consists of 13 sections and contains tables setting forth election statistics and election practices from each State. The Report provides a national overview of election law and practice, and provides a platform from which to examine New Jersey election law.

This memorandum sets forth summaries of select issues taken from the Report and other sources: (1) New Jersey voting patterns, (2) correlation between voter turnout and voter registration (3) statewide voter registration, (4) early voting, liberalized absentee voting, and vote by mail, and (5) effect of law on voter turnout. In addition, the laws of several states with progressive legislation are examined and summarized to provide greater detail where necessary.

Summaries of Conclusions

1. The most vital area for legal reform is the creation of a statewide voter registration system. It is the bedrock of election law and practice. A central list allows election administrators to: (a) coordinate voter registration with other voter registration agencies, such as the Division of Motor Vehicles, (b) eliminate duplicate registrations, (c) permit same day registration and (d) allow voters to vote from any site within the State.
2. Legislation that increases the cost of voting, such as registration requirements, depresses voter turnout. Legislation making it more convenient to vote may not appreciably increase voter turnout, but, where enacted, it is used and strongly supported by voters. Educated voters are more likely to use alternative voting procedures than less educated voters.
3. Early voting, liberalized absentee voting and vote by mail have not been around long enough to quantify whether the alternatives increase voter turnout, though modestly, or whether the modest increase in voter turnout is due to the novelty of the practice.
4. With the exception of statewide voter registration systems, the variety of state election laws and practices demonstrates that there is no “correct” legal regime for conducting elections. Individual state law is the product of local culture and historical accident.

New Jersey Voting Patterns

In New Jersey, the rate of voter turnout in the 1996 federal elections was 51.1% of the voting age population. The Report §2, 18. Over seventy (70.1) percent of persons eligible to vote were registered to vote in the 1996 federal elections and 86.4% of registered voters actually turned out and voted in the election. Id. at 14. The rate of voter turnout in the 2000 federal election was approximately the same, 50% percent.

Local elections fare far worse than statewide elections. In 1998, voter turnout for school board and budget elections was 14% of the eligible voting population. The New York Times, Metropolitan Desk April 23, 1998. The result prompted then Governor Christine Todd Whitman to call for moving school elections to November. That move has not happened.

New Jersey maintains voter registration lists at the county level. N.J.S.A. 19:31-6. To register, a person must be a U.S. citizen, at least 18 years of age, be a resident of the county for at least 30 days prior to the Election Day and not be on parole or probation or serving a prison sentence for any indictable offense. N.J.S.A. 19:31-5. The closing period for registration in New Jersey is 29 days prior to Election Day.[2]N.J.S.A. 19:31-6. New Jersey has absentee balloting for civilian and military service voters who invoke a statutory reason to request the ballot. N.J.S.A. 19:57-1 & 3. In 1996, federal election absentee ballots cast in New Jersey counted for 3% of the total vote. The Report, §5, 12.

New Jersey neither has a statewide voter registration system nor any form of alternative voting such as early voting, liberalized absentee balloting, vote by mail or Internet voting.[3]

Correlation between Voter Turnout and Voter Registration

The rate of voter turnout corresponds to the rate of voter registration. Jason P.W. Halperin, A Winner at the Polls: A Proposal for Mandatory Voter Registration, 3 NYU J. Legis. Pol’y 69 (1999). “Over the past twenty years, registered voters have turned out at a rate of 80 to 90 percent.” Id. at 72. Voter studies consistently show that even new voters, once registered, vote at very high rates. Id. The dominant reason for failure to register to vote is the burdens imposed by state registration laws. It follows logically - whether it follows empirically is a separate question - that if government wants to increase the rate of voter turnout, government must simplify the process of voter registration and consider the option of making voter registration mandatory.

Mr. Halperin uses the State of New Jersey in his case study to support his theses. Id. at 74. New Jersey reached its zenith of voter turnout in the 1840 presidential election when 80 percent of white adult males cast their vote. Since that time, voter turnout has been consistently downhill. Concerned with ineligible voters, the Legislature under Governor William Pennington enacted legislation to require any voter challenged at the polls to prove his citizenship. Id. at 75. In 1866, after the Civil War and after an influx of a substantial number of immigrants, the Legislature then enacted the first voter registration system in New Jersey administered by bi-partisan election laws. Id. at 76. This law, enacted by a Republican Legislature, was a reaction to unproved claims that Democrats had fielded ineligible voters from the cities to vote their party line. This landmark law shifted the responsibility of voter registration to individual voters. “Only those citizens who had taken the advance step of registering at some date prior to the election could vote.” Id. Before that time, a qualified citizen, able to prove citizenship if challenged, could walk into the poll and vote on Election Day without any prior registration.

In response to Democratic outcries that the voter registration law shut out the poor, the Republican Legislature in 1871 limited voter registration to seven cities with populations of more than 20,000 believed to be Democratic strongholds. Id. at 77. However, in 1876, the Legislature expanded registration to cover cities with a population of more than 10,000 and to allow registration by affidavit. In 1890, the Legislature enacted a reform law to establish new bipartisan boards in each district and to require election officials to conduct a house-to-house canvass of every voter. Id. at 78. Finally, Governor Woodrow Wilson, in response to unproved claims of fraud, convinced the Legislature to enact a law creating “a complete primary system, a secret ballot and vastly strengthened registration procedures.” Id.

The net result of this legislative activity over voter registration had one clear result: “New Jersey’s progression of stricter and stricter registration laws ultimately led to a substantial drop in voter turnout.” Id. at 79.

Statewide Voter Registration Systems

Voter registration is the first step in the election process in every state except North Dakota (which does not require registration to vote). The harder it is to register, the less likely it is that eligible persons will vote on Election Day. That correlation has been established since registration requirements were imposed after the Civil War. Decentralized registration systems suffer from list maintenance problems, impose burdens on voters who move and make it more difficult to coordinate voting lists between election officials and agencies authorized to register voters under the National Voter Registration Act in 1993.

The registration list is the foundation of state election practice. The evidence is clear. The only method to maintain accurate, current and reliable voter lists is to develop a statewide voter registration system. Statewide voter registration systems consist of a central database that contains the names, addresses and unique identification numbers of all registered voters in the state. Local election administrators have access to this database often in real time to register new voters, verify provisional ballots and eliminate duplicate registrations. Statewide voter registration systems require standardized data formats and computing platforms. To date, 20 states have statewide voter registration systems in place or in process.[4] Michigan is the largest state, 6.8 million citizens of voting age and the nation’s eighth largest electorate, to adopt statewide voter registration. Florida, with the fourth largest electorate of 9.8 million people, is expected to implement statewide voter registration.

Many factors have promoted the development of statewide voter registration systems: computerization of records, increased bandwidth and faster communication devices and innovations in network technology. However, the chief factor was the 1993 enactment of the National Voter Registration Act. That Act, better known as Motor-Voter, required states to allow voter registration through driver’s license bureaus and social service agencies. The Motor-Voter Act also set stricter standards for list maintenance and stricter standards for purging of the voting lists. The net result was swelling of the voter rolls and problems coordinating efforts between election registrars and other voter registration agencies.

Michigan enacted the Qualified Voter File (QVF) in 1993. Mi. St. 168.509m et seq. The Secretary of State was charged with the establishment and maintenance of the QVF, and the selection of the technology to implement the system. Mi. St. 168.509o. The various cities, townships, villages or school districts have electronic access to the QVF database. The Secretary of State manages individual mail-in registrations and instructs voter registration agencies to comply with the QVF requirements to ensure that there is only one official voting list in the State. The information required for individual voters is: (1) name, address, city, zip code and date of birth; (2) the person’s driver license number, state personal identification card number or similar number; (3) jurisdictional information; (4) precinct numbers and ward numbers, if any; (5) other information the Secretary of State requires; and (6) voting history for last five years. Mi. St. 168.509p. The Privacy Act of 1974 precludes the use of social security numbers as means of identifying voters. Consequently, as a practical matter, Michigan uses driver’s license numbers in most cases.

Michigan’s QVF represents the best model for legal reform in voter registration systems. Michigan is a large, industrial state fragmented into 273 city and 1,242 township governments. The State provided hardware and software to 83 county clerks and 236 clerks of cities and townships with a population over 5,000. Ninety-four smaller jurisdictions purchased the equipment at their own expense. The system has saved local government the cost of voter list management. Since New Jersey, unlike Michigan, has only 21 county registration offices to unify, the transition to a statewide voter registration system in New Jersey would be easier and less costly to establish while at the same time yielding the undisputed benefits of accurate voter lists. N.J.S.A. 19:31-6 (registration lists maintained at county level).

Statewide registration lists also enable the following: (1) facilitation of statewide provisional balloting since election officials can quickly verify registration in other counties, (2) same day registration since persons would be unable to vote from different sites, (3) (coupled with electronic voting systems) for registered voters to vote anywhere in the State. Nationwide, as of 1990, 17.7% of employed persons worked outside the county of their residence and 30% worked more than 30 minutes away from their workplace.

Early Voting, Liberalized Absentee and Vote by Mail

In the last decade, states have adopted a number of measures to allow citizens to vote on a day other than Election Day.[5]The Report, §5, 1. Many states, especially in the West, have followed the lead of California in 1978 and liberalized access to absentee ballots. Altogether, 22 states now make an absentee ballot available to any registered voter who requests one, without need to show cause. Thirty-two percent of the voting age citizen population lives in a state that provides an absentee ballot automatically upon request.

Fourteen other states have adopted "early voting" procedures.[6] Pioneered by Texas in 1991, early voting evolved from in-person absentee voting, but is now distinct from it. In-person absentee voters must apply for an absentee ballot; early voters must simply report to an early voting station, sign the poll-book, and have registration verified. Absentee ballots cast in person are usually enclosed in a sealed and signed envelope; early voting ballots cannot be identified individually. In-person absentee voting, finally, takes place only within the regular hours of the elections office; early voting programs often provide extended hours on both weekdays and weekends. Twenty-five percent of the voting age citizen population lives in states with early voting.

Finally, in 1995, Oregon became the first state to implement voting by mail (VBM) in statewide elections, employing it first in special partisan primary elections and soon after in a special general election to fill a vacant United States Senate seat. In 1998, by more than a two-to-one margin, Oregon voters approved an initiative to extend vote by mail permanently to statewide primary and general elections.

Early voting programs overlap significantly with liberalized absentee laws. All but four of the states that provide early voting also provide absentee ballots automatically upon request. Taking the three provisions together, 26 states with 44.5 percent of the voting age citizen population make it easy for voters to cast their ballots before Election Day. State courts in Oregon, Tennessee and Texas have upheld those states’ early voting procedures. The United States Supreme Court rejected an appeal of the Texas law. The United States District Court for the Eastern District of North Carolina Western Division dismissed a challenge to that state’s no excuse absentee voting system.

The effect of early voting and liberalized absentee voting has been dramatic. In 1980, just as the movement toward liberalization of access to absentee ballots was begun, five percent of voters nationwide cast their votes by absentee ballot. In 1996, 10 percent nationwide voted prior to Election Day, either by mail, mostly by absentee voting, or in-person before Election Day. In states with liberal access to voting before Election Day, the percentages are still higher. Thirty-nine percent of the 2000 presidential vote in Texas was cast early, and 24.6 percent of the 2000 vote in California was by absentee ballot. In Oregon, every statewide election since 1995 has used a mail-in ballot.

The stated rationale for the extension of early voting, vote by mail, and the liberalization of access to absentee voting was to make it easier for people to vote. Studies of voter turnout and surveys of non-voters both have found that large numbers of people do not vote because it is costly or inconvenient for them. By making it easier to vote, it was argued, participation in elections would increase.

However, measured against the standard of promoting turnout, the innovations have been either a modest success or a modest disappointment, depending upon expectations. The consensus among analysts is that liberalized absentee voting has had a very small positive effect on voter participation. The most careful study of the effect of liberalized absentee voting, by Eric Oliver of Princeton University, found increases in voter turnout in states that had liberalized. But the increases in turnout depended wholly upon the ability of political parties to mobilize voters. In states with closed primaries, and therefore with registration of voters by partisanship, turnout increased modestly, by just over 2 percent. In states with open primaries, and therefore no easy identification of voters by partisanship, turnout increased far less if at all. Liberalized access to absentee balloting generally increases voter turnout by perhaps a small amount.

Studies of early voting have tended to find larger but still modest effects on voter participation. A study of Tennessee found a five percentage point increase in turnout in the first early voting election, 1994, compared to nine earlier midterm congressional election years. Another study of Texas found that increases in levels of early voting across counties in 1992 correlated positively with increases in voter turnout. These studies suggest a relationship between early voting and voter turnout. But it is hard to tell whether increases in voter turnout are the permanent result of greater ease of early voting or are the temporary result of the novelty of early voting. Moreover, it is difficult to assess whether the increases in turnout that analysts have seen in aggregate turnout figures are the result of early voting or the result of other steps that states have taken simultaneously to promote voter turnout (e.g., easing access to voter registration or more strongly encouraging people to vote in publicizing early voting) or the result of other features of the counties or elections that were compared (e.g., more competitive elections coincident with the introduction of early voting or political characteristics of counties that cause them to have higher rates of early voting and higher rates of voter turnout).