Testimony of Justin Levitt before the Wisconsin State Legislature

March 31, 2010

Click here to download this resource as a PDF.

Testimony of

Justin Levitt, Counsel
Brennan
Center for Justice at NYU School of Law

Before the Wisconsin Assembly Committee on Elections and Campaign Reform
and the Wisconsin Senate Committee on Labor, Elections and Urban Affairs

March 31, 2010

Good morning, distinguished Members.  My wife’s family is from Madison, and so I am particularly pleased to be back here; I thank you not only for the chance to come back into town, but also for the opportunity to speak at this very important hearing.  My name is Justin Levitt, and I am an attorney at the Brennan Center for Justice at NYU School of Law.  The Brennan Center is a non-partisan organization that unites scholars and advocates in pursuit of a vision of inclusive and effective democracy.  Toward that end, the Center’s Democracy Program researches and promotes reforms that eliminate barriers to full and equal political participation and that foster responsive and responsible governance.  

In particular, the Brennan Center has been deeply involved in the effort to ensure fair and accurate voter registration systems and thereby promote eligible citizens’ participation in elections, both in the Midwest and across the country.  We have done extensive work on the building and maintenance of voter registration lists and on practices that enhance and detract from the ability of those lists to reflect the electorate.  This work has included thorough research, assistance in drafting regulatory and legislative provisions, testimony before bodies like these honorable committees, and, where necessary, litigation to compel compliance with obligations under state and federal law, including the most fundamental guarantees of our Constitution.

Most recently, we have been among the national leaders in encouraging procedures that modernize our voter registration system.  In addition to our direct work with state and federal officials, we have produced at least three major reports and six further briefing papers chronicling the need for modernization, the accuracy and efficiency savings that would result, and the availability of models to point the way.  I am therefore particularly delighted to testify before you today, as you consider AB 895 and 892, and their companion bills SB 640 and 645.  We urge the Committee to ensure that meaningful action on the bills follows in a manner as timely as these hearings.

Today, I hope to briefly lay out an overview of our conclusions on the need for reform and the components that we believe an effective modernization program should incorporate, drawing from the publications that I mentioned above.  In most respects, the bills before you today embrace the fundamental reforms we urge, and many Wisconsinites would benefit from their passage.  Given our experience with the topic, however, we do have a few suggestions that we hope you consider as the bills move forward.  Most of these recommendations are small, but a few are larger, scaled to the admirable ambition of the modernization project.

The need for an upgrade

Our democracy is a source of pride and strength.  But while we have regularly upgraded many components of our election system to keep pace with both American technology and American demography, the voter registration system stands out as a significant anomaly, rooted firmly in nineteenth-century procedures.  The system is prone to error and manipulation, and includes substantial gaps.  Nearly a third of voting-age citizens nationwide, for example, are not registered.[1]  Election officials, in turn, face an expensive and inefficient crush of new registrants every election cycle.  It is time for an upgrade.

To your credit, and to the lasting credit of those who administer your elections, many registration practices are more effective in Wisconsin than in many states.  Even so, Wisconsinites could be substantially better served by a modernized registration system.  In November 2004, the Census Bureau estimated that 18% of Wisconsin’s four million voting-age citizens were not registered to vote.[2]  By November 2006, that percentage had increased to 28%.[3]  And though the November 2008 electorate was described by many as newly energized, still 26% of Wisconsin’s voting-age citizens were unregistered.  By these estimates, one million voting-age citizens — more than one in four — are not registered to vote in the state.[4]

In part, this is because the voter registration system is simply not designed for a modern mobile society.  One in six Ameri­cans moves in a year, and about 45% of the eligible voting population moves in any one five-year period.[5]  The Wisconsin electorate reflects similarly high mobility: approximately 13% of the voting-age population moves within the state each year.[6]  At this rate, even when people become registered, they do not stay registered.  

The current system is also prone to error, which can lead to disenfranchisement. In our work, we have extensively studied attempts to match voter registration records to other data sources, which I know is a topic of particular salience here in Wisconsin.[7]  It is not unusual to find an initial error rate of 15-30%, often driven by glitches as simple as typographical errors or inconsistent representation of compound names in the voter rolls.[8]  These mistakes or inconsistencies, which can make it impossible to verify a registration or to find a name on the pollbooks, are the inevitable product of a system based on paper forms that arrive in an increasing flood as election day approaches.

The voter-initiated paper registration system also invites manipulation or the fear of manipulation.  Isolated reports of irregularities have spurred increased efforts to impose new restrictions on voter registration, with burdens borne at the end of the day by eligible voters.  This includes cumbersome restrictions on voter registration drives, of the sort that recently shut down registration efforts by Florida’s League of Women Voters for the first time in 70 years.[9] They also extend to voters blocked from the rolls by minor technical errors or purged from the rolls by overly sweeping attempts to find the ineligible, or unduly harsh documentation requirements for registration that otherwise qualified registrants find difficult to meet. A modernized registration system can increase everyone’s confidence in the process without undue measures that disenfranchise real voters.

It is also important to remember that the limits of the existing system do not fall equally on all Americans. Registration problems of the present have a disproportionate impact on low-income citizens and those who are less educated. Such individuals are more likely to move more often with a need to re-register upon every move, to have an unconventional living situation that makes verification of eligibility difficult, and to have limited access to the information and forms necessary to register and stay registered.  In part as a result, the Census Bureau estimates that voting-age citizens without a high school diploma are left unregistered almost twice as often as eligible voters with a high school diploma or above, and more than three times as often as eligible voters with an advanced degree; voting-age citizens from households making less than $30,000 per year are left unregistered twice as often as those making more than $100,000 per year.[10]    

In many ways, Wisconsin’s election-day registration system is a secure and valuable failsafe mechanism that mitigates some of the damage above.  It does not, however, resolve an important set of problems related to registration before election day.  Citizens who are not on the rolls — or not on the rolls in the right place — will not receive a sample ballot before an election, or a mailing with the location of their polling place, or other official notice from the state that an election is imminent. They are far less likely to receive mailings from candidates and to be canvassed by volunteers. They will not usually be called by pollsters or contacted by nonpartisan groups doing voter education. In short, because personal voter contact is driven largely by the registration list, those who are not on the rolls will receive much less of the individualized contact that has been shown to spur turnout.[11]

Election-day registration also does not completely capture substantial time and cost efficiencies that modernization would provide.  As mentioned above, a voter-initiated system of handwritten forms fosters mistakes, including inadvertent omissions and multiple duplicate forms with the same information.  Citizens may register with a nickname or married name that is different from the name that appears in other government records, making the data difficult to verify.  Their handwriting may be difficult to read, leading to multiple rounds of typos and corrections.  And particularly as election day draws closer, increasing publicity about the election process may spur multiple redundant or unnecessary attempts to register, with the same mistakes or new ones. 

Officials valiantly attempt to correct or compensate for these errors, but the corrections take time, and often, extra staff.  Moreover, large numbers of forms are invariably submitted in a last minute rush that is wholly predictable — the IRS estimates that more than 20% of taxpayers wait until the last minute to file their taxes[12] — but this too strains official resources, with the need to provide temp time and training.  It also increases the difficulty of making rational decisions about allocating the necessary ballots, privacy booths, and poll workers.  Long lines and disgruntled or disenfranchised voters result.

The elements of an upgrade

In order to ensure that all eligible citizens are registered without the gaps and inefficiencies mentioned above, the Brennan Center has consistently emphasized four primary components for bringing a voter registration system into the 21st century: automation, permanence, online access, and an election-day failsafe. 

Automated registration extends existing tools and resources to ensure that all eligible citizens are registered to vote, without the need to submit redundant paper forms.  Many state agencies, in the course of their regular data collection, gather reliable records sufficient to identify eligible voting-age citizen residents.[13]  These may include motor vehicle agencies, public assistance agencies, disability agencies, tax authorities, and educational agencies, among others.  Automated registration structures use these existing records as a basis for identifying voters, without unnecessarily recreating the wheel.  They forward data records electronically to election authorities, who can compare the records to the existing statewide voter rolls, and add new eligible voters as a default matter.  The data transfer process need not supplant any list maintenance or double-check procedures that election officials now undertake whenever they receive a paper registration form, and citizens may opt-out of the registration process at any time.  In all, the automation increases accuracy and security and efficiency: the need for costly data entry decreases, the reliability of the information increases, and the flow of registration records is flattened and spread across the entire election cycle rather than concentrated in the final push to election day.

Permanent registration ensures that voters, once entered onto the state rolls, stay registered so long as they continue to live in the state, without the need to constantly update registration information if they move or if their circumstances change.  Using the same reliable data described above, transferred electronically to maximize accuracy and efficiency, a voter updating her name or address at a cooperating state agency would have the same information reflected on the voter rolls, keeping the rolls consistently up to date.  This too cuts down on the need for multiple redundant forms, and the time and confusion at the polls that results when voters update one reliable government record but forget to update their voter registration.

Online access to registration ensures that the voters who know best are able to find, verify, and update their own information online — and that they are able to submit a new request for registration if they are not already reflected on the rolls.  The single best source of up-to-date, accurate voter information is the voter herself.  Ensuring that a voter is able to securely double-check her own registration status and data, and efficiently communicate changes or mistakes, is the best way to insure that the rolls remain accurate.

Finally, we encourage an election-day failsafe mechanism, for eligible citizens who somehow slip through the cracks to correct omissions or other problems on election day itself.  The more that the remainder of the registration system is modernized, the less that these election-day correction procedures will be necessary.  Still, in order to ensure that eligible electors are not unduly turned away from an election, some failsafe procedure will always be necessary.

Wisconsin’s system and the bills before the Committees

Much of the above description, of course, likely sounds familiar to you.  Wisconsin was one of the earliest adopters of election-day registration; studies indicate that the policy has yielded great success in boosting turnout while maintaining security, and in limiting the cost and uncertainty of provisional ballots.[14]  Wisconsin has also implemented a system for voters to check their registration information online, though the system does not yet permit citizens to directly request a new registration or correct information that is mistaken or outdated.[15]  These are essential elements of a modern voter registration system, and they now benefit eligible Wisconsinites in every cycle.

The bills before you today represent carefully structured efforts to make up much of the rest of the ground, leading a wave of modernization efforts, large and small, across the country.[16]  AB 895 and SB 640, of course, have multiple provisions, including several related only tangentially to voter registration — but most relevant to my remarks today, they would dramatically increase the degree to which motor vehicle records are used to build a system of automated and permanent registration.  AB 892 and SB 645, in turn, would expand Wisconsin’s existing online system, allowing a citizen not only to look up her registration record, but to submit a secure registration request or update online as well.  Both sets of bills would increase the accuracy and efficiency of Wisconsin’s voter registration system, and in line with the experience of similar automated systems elsewhere, we would expect a dramatic decrease in registration costs.[17]  These are worthy goals benefiting Wisconsin officials, Wisconsin voters, and Wisconsin taxpayers alike, and we support them.

At the outset, however, I promised recommendations to build on the measures before you, to help improve them as they pass through the legislative process.  Our single biggest concern is that both sets of bills are keyed to the Department of Transportation’s records, which are themselves largely limited to Wisconsinites with a driver’s license or non-driver’s state ID.  This threatens to leave behind a sizable population of eligible Wisconsin citizens; researchers in 2005, for example, found that hundreds of thousands of voting-age Wisconsinites did not have a driver’s license or non-driver’s state ID.[18]  Moreover, the demographics of this population reflect historically underserved communities that should instead merit concerted efforts for inclusion: those without valid DOT identification (and presumably valid DOT records) are disproportionately elderly or 18-24 year-old minorities.[19]

At present, AB 895 and SB 640 at least ask for an evaluation of state agencies beyond the DOT that may better capture the eligible citizens not represented in DOT systems.[20]  This is welcome, but insufficient, for the same reasons that federal voter registration reforms have not focused on motor vehicle agencies alone.  We urge the committee to consider more binding commitments to modernized registration drawn from other reliable state sources; one such mechanism might require inclusion of additional state agencies upon certification by the Government Accountability Board that the source agency’s data and transmission infrastructure were sufficiently reliable to meaningfully promote the completeness and accuracy of the statewide voter registration list.

We would also offer other suggestions to improve the bills before you.  At present, for example, when data records from the Department of Transportation reveal citizens not already on the voter rolls, the bills ask the Government Accountability Board to seek any additional information necessary to register those would-be voters.[21]  We would also suggest granting the Department of Transportation authority and encouragement to request that information — which will usually amount to no more than two or three incremental questions — up front during the licensing process.  We find that this model, already in place in states like Delaware, Florida, and Kansas, yields the greatest efficiencies in the registration process.[22]

Similarly, we recommend expanding the information conveyed from the Department of Transportation to the Government Accountability Board,[23] to include name history and at least partial Social Security number digits.  Such information can help avoid some of the predictable errors in matching information from system to system, ensuring that duplicate voter records are identified and that voter information is reliably updated even after a change of name or the acquisition of a new driver’s license.

We might also suggest, for example, minor revisions to the portion of the bills concerning the matching of information from system to system in order to verify the voter registration information and comply with federal law.  At present, under certain circumstances, the bills essentially require checking voter registration information that has been gleaned from DOT records against DOT records themselves[24] — akin to checking a reliable photocopy against an original document.  The language can be modified slightly to remove the unnecessary redundancy in a way that will neither impede smooth registration nor depart from applicable federal requirements.

There are a few other more minor suggestions that we would be pleased to provide separately, without taking the Committees’ time today.  Likewise, I have restricted my remarks today to the portions of the bills before you that work to modernize Wisconsin’s voter registration system.  There are other elements of these bills that we support, including provisions to bring Wisconsin into compliance with the MOVE Act, provisions to address some of the deceptive and intimidating practices that have heretofore escaped meaningful redress, provisions to create and post a voter’s bill of essential rights, and provisions designed to increase the meaningful efficacy of challenges and reduce their overbroad exercise, among others.  The Brennan Center has extensively studied, and occasionally litigated, many of the policies behind these provisions, and we similarly have a few minor recommendations which we would be pleased to provide under separate cover.

Overall, however, I want to again commend the Committees for the exceedingly important work you are undertaking in working to modernize Wisconsin’s voter registration system.  Despite the successes that Wisconsin has already seen, there are many more eligible citizens to bring into the active electorate through modernization, and many more efficiencies to be gained at the same time that the accuracy of the system overall is substantially enhanced.  These bills that you are considering reflect a broad mandate to bring the voter registration system into the 21st century, which we heartily applaud.  And as much as we welcome the effective use of Department of Transportation data records to improve voter registration in Wisconsin, we hope that the current bills can be expanded to embrace in decidedly more robust fashion agencies beyond the Department of Transportation as well.

Thank you again for the opportunity to speak with you today.  As with other states across the country, the Brennan Center stands ready to assist you and the whole of Wisconsin in your efforts to make the election process as effective as it can possibly be.  I thank you very much for your time – and I am more than happy to answer any questions that you may have.


[1] U.S. Census Bureau, Voting and Registration, Voting and Registration, Reported Voting and Registration of the Citizen Voting-Age Population, for States: November 2008 tbl. 4a [“2008 Census”], available at http://www.census.gov/hhes/www/socdemo/voting/publications/p20/2008/tables.html.

[2] U.S. Census Bureau, Voting and Registration, Voting and Registration, Reported Voting and Registration of the Citizen Voting-Age Population, for States: November 2004 tbl. 4c, available at http://www.census.gov/hhes/www/socdemo/voting/publications/p20/2004/tables.html.  The Election Assistance Commission has also produced estimates of the registration rate for various states.  See, e.g., U.S. Election Assistance Commission, The Impact of the National Voter Registration Act of 1993 on the Administration of Elections for Federal Office 2007-2008, at 27 tbl. 1a & nn. (2009).  Census Bureau estimates are based on individual self-reporting of registration status; the biannual EAC reports reflect a lower unregistered percentage, but are based on jurisdictions’ aggregate reporting of registration volumes, and are likely to include duplicates and/or individuals reflected on the rolls who have moved out of the state or passed away.

[3] U.S. Census Bureau, Voting and Registration, Voting and Registration, Reported Voting and Registration of the Citizen Voting-Age Population, for States: November 2006 tbl 4a, available at http://www.census.gov/hhes/www/socdemo/voting/publications/p20/2006/tables.html.

[4] 2008 Census, supra note 1, tbl. 4a.

[5] See Haya El Nasser & Paul Overberg, Millions More Americans Move to New States, USA Today, Nov. 30, 2007; see also U.S. Census Bureau, American Factfinder: Geographic Mobility by Selected Characteristics in the United States — United States, tbl. S0701, available at http://bit.ly/aSbu3h; Voter Registration: Assessing Current Problems: Hearing Before the S. Comm. on Rules & Admin., 111th Cong. (Mar. 11, 2009) (statement of Nathaniel Persily), available at http://bit.ly/aqk5Ty.

[6] U.S. Census Bureau, American Factfinder: Geographic Mobility by Selected Characteristics in the United States — Wisconsin, tbl. S0701, available at http://bit.ly/9zUx1D.

[7] See, e.g., van Hollen v. Gov’t Accountability Board, Case No. 08-CV-4085 (Wis. Cir. Ct. 2008).

[8] See, e.g., Decl. of Andrew Borthwick ¶¶ 20-51, Florida State Conference of the NAACP v. Browning, Case No. 4:07-cv-402 (N.D. Fla. Sept. 17, 2007).

[9] See, e.g., League of Women Voters v. Cobb, 447 F. Supp. 2d 1314, 1325-26 (S.D. Fla. 2006).

[10] 2008 Census, supra note 1, tbls. 6 & 8.

[11] See generally Donald P. Green & Alan S. Gerber, Get Out the Vote: How to Increase Voter Turnout (2d ed. 2008).

[12] U.S. Internal Revenue Service, IRS Urges Taxpayers to Avoid Last Minute Filing, IR-2003-40, Mar. 27, 2003, available at http://www.irs.gov/newsroom/article/0,,id=108441,00.html.

[13] See Laura Seago, Government Lists: How Ready Are They for Automatic Registration? (2009), at http://bit.ly/9lcwdt.

[14] See, e.g., Barry C. Burden et al., The Effects and Costs of Early Voting, Election Day Registration, and Same Day Registration in the 2008 Elections (2009), at http://electionadmin.wisc.edu/pewreportfinal.pdf; DÄ“mos, Voters Win with Election Day Registration 5 (2009), at http://www.demos.org/pubs/voterswin_09.pdf; Lorraine Minnite, Election Day Registration: A Study of Voter Fraud Allegations and Findings on Voter Roll Security (2007), at http://www.demos.org/pubs/edr_fraud_v2.pdf; Benjamin Highton, Voter Registration and Turnout in the United States, 2 Perspectives on Pol. 507, 509 (2004).

[15] See State of Wisconsin Voter Public Access, at https://vpa.wi.gov.

[16] See Wendy Weiser, Christopher Ponoroff & Nhu-Y Ngo, Modernizing Voter Registration: Momentum in the States (2010), at http://bit.ly/cRIawQ.

[17] See id. at 4.

[18] See Univ. of Wisconsin-Milwaukee, Voter Photo ID Law Court Cases Utilize ETI Research (2007), at http://www4.uwm.edu/eti/2007/VoterID.htm; John Pawasarat, The Driver License Status of the Voting Age Population in Wisconsin (2005), available at http://www.uwm.edu/Dept/ETI/barriers/DriversLicense.pdf.  The Department of Transportation has published more recent statistics on the total coverage of DOT licensees and photo identification card holders, but these statistics do not indicate which individuals still reside within the state.  See, e.g., Wis. Dep’t of Motor Vehicles, Facts & Figures 2008, at 43, 54-56, available at http://www.dot.wisconsin.gov/drivers/docs/2008ff.pdf

[19] See Pawasarat, supra note 18.

[20] A.B. 895, § 132(4).

[21] A.B. 895, § 30 (creating Wis. Stat. § 6.256(4)).

[22] See, e.g., Wendy Weiser & Christopher Ponoroff, Paperless Voter Registration (forthcoming 2010).

[23] A.B. 895, § 30 (creating Wis. Stat. § 6.256(2)).

[24] A.B. 895, §§ 42, 46 (creating Wis. Stat. §§ 6.34(2n), 6.34(4)).