Online voter registration in Oregon: towards an election administration triple bottom line.
If hanging chads and provisional ballots were the top election administration controversies of the 2000 and 2004 elections, then voter registration was undoubtedly the defining issue of the historic 2008 election. (1) States were flooded with more than 60 million voter registration applications between 2006 and 2008. (2) Legal clashes over the actions of registration drives and purging practices of statewide voter registration databases saw progressives decrying unnecessarily restrictive policies that could result in eligible registrants left off the rolls, while conservative groups invoked the specter of fraudulently obtained registrations. (3)
At the center of this massive logistical challenge and rash of bitterly contested litigation was the humble paper voter registration card, delivered by the millions to county elections offices to be processed, transcribed, and matched to driver's license or social security records. The system of paper-based voter registration is both enormously expensive and prone to human error at every step along the way: registrants commonly omit required information and write illegibly, and county elections offices inevitably commit transcription errors. (4) Millions of voter registrations that can't be matched to government records are the natural result, which threatens the right of qualified registrants to cast a ballot on Election Day and fuels fears of fraudulent voter registration practices.
Partisan conflict over election administration practices shows no sign of abating. But online voter registration, a recent election administration policy innovation pioneered by state legislatures, promises to bridge the partisan divide and unite election reformers, budget hawks, and voter fraud watchers around an election administration triple bottom line (5): increasing participation, decreasing costs, and improving integrity.
Online voter registration went live in Arizona by 2002, but it wasn't until 2008 that the state of Washington had joined Arizona in rolling out a system of its own. By the end of 2009, Oregon, Colorado, Indiana, Kansas, Louisiana, Utah, and California had passed online voter registration legislation, and at least four other state legislatures had introduced bills proposing similar systems. (6) While online voter registration has attracted press coverage and praise from commentators across the political spectrum, (7) it has received limited analysis by academics and policy experts. (8) This Note focuses on online voter registration legislation in Oregon as a case study of the second wave of states to adopt such a system. (9)
Online voter registration enhances Oregon's traditional, paper-based voter registration system in three primary ways. First, online registration lowers barriers to registration and decreases errors in the registration process. In these respects, Oregon House Bill 2386 aimed to boost overall rates of registration and turnout among eligible voters generally, and young voters in particular.
Second, online voter registration eliminates the need for costly data entry and processing of paper voter registration cards, yielding considerable savings to cash-strapped state and local governments.
Third, when combined with well-designed statewide voter registration databases, (10) electronic data transaction standards, and eligibility verification processes, online voter registration further reduces the already low risk of voter registration fraud and helps ensure the integrity of ballots cast on election day.
I. BROAD SUPPORT FOR OREGON'S ONLINE VOTER REGISTRATION BILL
Oregon's reputation as an innovator in election administration stems from its first-in-the-nation adoption of a statewide vote-by-mail system, credited with increasing turnout among registered voters, reducing costs, and increasing election security. (11) But vote-by-mail's requirement that a voter have her current residence address on file in order to receive a ballot negatively affects the participation of highly mobile populations: young people, low income people, recent immigrants, and communities of color, in particular. (12)
Such concerns weighed upon thirty-one-year-old Representative Ben Cannon, who drafted and introduced Oregon House Bill 2386 (HB 2386) after receiving an email from a young college student asking why Oregon hadn't followed the lead of Washington and Arizona in establishing an online voter registration system. (13)
A coalition of online voter registration stakeholders coalesced early in the legislative session. Supporters included Cannon, nationally recognized youth-vote advocate Representative Jefferson Smith, key budget negotiator Senator Richard Devlin, good-government organizations like the League of Women Voters, youth-vote organizations like the Oregon Student Association and the Oregon Bus Project, and newly elected Secretary of State Kate Brown. Finally, the coalition benefited from the enthusiastic support of an unlikely partner: Representative Sal Esquivel, a self-described conservative and cosponsor of a controversial voter ID bill. (14)
HB 2386 directs the Secretary of State to adopt by rule an electronic voter registration system, and outlines a few specific requirements for the system. Most significantly, the bill mandates that only those who possess a valid Oregon driver's license, driver's permit, or state identification card may use the online voter registration system. (15) The bill further instructs the state Department of Transportation to provide the Secretary of State with a digital copy of the relevant state ID for every person who submits her registration online. (16) Since online registrants can't "sign" their registration online, county elections offices need a digital image of registrants' signatures from their state IDs to verify the signature on the voters' ballot envelopes.
In the middle of the worst recession in a generation, even the most worthy new initiative that depended on state funds would have been dead on arrival. However, the bill's advocates were able to draw entirely upon federal funds appropriated under the federal Help America Vote Act (HAVA) for the upgrading of state electoral systems. The Secretary of State budgeted one-time costs of $200,000 for the development of the voter registration system itself, and $17,000 for the development of a mechanism to transfer registrants' digital signatures from the DMV. During the first budget cycle in which the system will be fully operational, the Secretary of State projected $22,000 in operating costs. (17)
HB 2386 underwent minor technical amendments and faced no organized opposition, passing the House and Senate with broad--but not unanimous--bipartisan majorities, (18) and Governor Ted Kulongoski signed it into law in August 2009. Once the specifications for the system are settled through administrative rulemaking, (19) the Secretary of State will put the project out to bid, and the system is scheduled to go live by March 2010.
II. ONLINE VOTER REGISTRATION ACHIEVES THE ELECTION ADMINISTRATION TRIPLE BOTTOM LINE: ACCESS, EFFICIENCY, INTEGRITY
A. Online Voter Registration Increases Participation
It's no secret that the number of voting-eligible American citizens far outstrips the number of citizens who participate in the electoral process by voting. In Oregon, although turnout is relatively high among registered voters, the fact remains that more than twenty percent of eligible Oregonians are not registered to vote and thus do not participate. (20)
The causes of American under-participation are many and varied; the modern requirement that citizens register to vote is undoubtedly one. (21) Voter registration requirements not only prevent eligible but unregistered voters from participating, but they can also prevent registered voters from casting a ballot. One recent nationwide survey found that registration problems were either a major or minor factor in preventing at least twenty percent of registered non-voters from participating in the 2008 election, accounting for roughly 2.2 million lost votes. (22)
From an expanding access perspective, problems with voter registration policies and procedures can be generally divided into two major categories: those that make eligible citizens less likely to register to vote and participate in the first place, and those that make citizens who actually submit a voter registration form less likely to cast a ballot on election day. Online voter registration addresses both the "front end" and the "back end" in significant, but limited ways.
1. Online Voter Registration Reduces Barriers to Registration
In contrast to many other industrialized democracies, the act of registering to vote in the United States is voluntary and the responsibility of the voter. (23) Such a diffusion of responsibility erects barriers to registration beyond voters' simple lack of desire to register.
One major barrier is the number of times a voter must register to vote. Since voting is tied to place of residence, more transient citizens--who tend to be younger and poorer--have to fill out voter registration cards more often, which entails more trips to the library, election office, or DMV. (24)
Another barrier is knowledge and experience in dealing with voter registration practices and government bureaucracies in general. (25) Many registered voters, especially in Oregon, don't know that because the post office cannot forward mailed ballots, voters must update their registration every time they move. (26) Confusing rules and instructions as well as unpredictable international mail delivery lowers the registration rates of overseas and active duty military service members. (27) Finally, perhaps as many as twenty percent of young would-be voters simply miss voter registration deadlines. (28)
Online voter registration can significantly mitigate the impact of these and other barriers to voter registration by reducing the cost and number of steps necessary to register to vote. Currently, registrants must pick up or download and print a voter registration form, fill it out, and then stamp and mail it or drop it off at an elections office. With an online registration system, an eligible registrant with Internet access anywhere in the world can initiate and complete the entire voter registration process in just three to four minutes, without buying stamps, gas, or bus fare.
Online voter registration is well suited to the human tendency to register at the last possible minute. Late registrants tend to be disproportionately young and geographically mobile, and can constitute anywhere from five to twelve percent of the electorate in a given election. (29) By reducing the number of steps that need to be completed when it matters most--the final days and hours before the deadline--online voter registration makes it more likely that last minute registrants will have a chance to participate.
Online voter registration also presents a unique opportunity to reach millions of potential young voters in a familiar web-based medium, (30) facilitated through online social networks. (31) Online voter registration will enable an Oregonian Facebook user to go from browsing her news feed to submitting her voter registration application in just a few clicks. (32)
While experts agree that adopting practices that make it easier and more convenient for citizens to register to vote has a significant positive impact on participation, (33) such voter registration practices--including online voter registration--have significant limitations.
Eligibility is the most obvious limitation. In order to register to vote online in Oregon and other states, the registrant must possess valid state-issued identification, which leaves out tens of thousands of eligible registrants who recently moved to the state or do not possess a state ID. The digital divide is another limitation. Differential rates of internet access between white registrants and registrants of color, as well as between registrants of different socioeconomic statuses, leave open the possibility that rather than broadening participation, online voter registration will merely make it easier for already high-propensity voters to participate. (34)
Indeed, the adoption of an online voter registration system does not alter the fact that younger, less affluent, and more geographically mobile voters disproportionately bear the burdens of frequent re-registration. Nor does the adoption of online voter registration alter voter registration deadlines or necessarily increase civic knowledge or engagement.
In short, while the effect of online voter registration on turnout will be significant, (35) the practice is no panacea to the problem of under-participation. Nor is online voter registration a replacement for other election reform efforts aimed at boosting turnout, such as same-day registration and automatic registration.
2. Online Voter Registration Reduces Errors in Registration
The most dramatic effect of online voter registration on voter participation may lie not in bringing more registrants through the "front end" of the voter registration system, but rather in strengthening the system's "back end": ensuring that citizens who submit a voter registration are actually able to cast a ballot on election day.
While problems with voting machines have drawn much media attention and generated some of the highest-profile litigation, (36) a far greater number of votes are "lost" before voters attempting to participate even have a chance to pull the lever (37)--or, in Oregon, to lick the envelope. Nationally, roughly 2.2 million votes were lost in 2008 because of registration problems--compared to 2.2 million lost because of voters' lack of identification, and 1.9 million lost because voters couldn't find their polling place. (38) While registration problems may be relatively less pronounced in Oregon, (39) the fact remains that a significant percentage of ballots are returned to county elections offices marked undeliverable. (40)
Though the biggest share of these undeliverable ballots likely results from registered voters who move and forget to re-register, a significant proportion of undeliverable ballots comes from simple human error on the part of the registrant (omitting necessary information, writing illegibly) or on the part of county elections offices (committing errors in manual data entry). (41) While some degree of human error in written communication is inevitable in any voter registration system, online voter registration can dramatically reduce its prevalence.
Online voter registration systems do not permit registrants to submit their registration form without completing all required fields, and eliminate errors based on illegible handwriting. Most significantly, online voter registration obviates the need for expensive and error-ridden manual data entry currently performed by county elections officials and temp workers.
Some of the very strengths of online voter registration relative to paper forms--web accessibility and centralization--are also weaknesses. A computer crash or web slowdown, or worse, a denial of service attack, in the days or hours before a registration deadline could potentially disenfranchise thousands of would-be registrants relying on the system. (42) An investment in fail-safe backup servers could pay substantial dividends in mitigating the effects of a crash and ensuring public confidence in the system.
B. Online Voter Registration Reduces Election Administration Costs
Voter registration is an expensive undertaking for state and county governments. One authoritative study pegged the cost of each voter registration transaction in Oregon at $8.43--a per-unit price that adds up quickly over the course of over a million such transactions in 2008 alone. (43) By drastically reducing the need for printing, distributing, and manually processing paper voter registration cards, online voter registration can dramatically cut costs even as it increases accessibility and integrity. For example, Maricopa County (Phoenix) reported at least an $0.80 reduction in processing costs for each voter registration application submitted online. (44) Similar figures on costs and usage in Oregon could have saved the state hundreds of thousands of dollars in the 2008 election cycle. (45)
However, implementing online voter registration requires hundreds of thousands of dollars in up-front investment to build the online interface, and the efficiencies it yields depend on factors difficult to forecast with precision, such as the rate of use over time. For these reasons, legislative budget analysts are unlikely to credit online voter registration legislation with net cost reductions over a given fiscal period. Federal HAVA funds specifically appropriated for states to upgrade voter registration systems are thus crucial to overcoming state budgetary limitations.
Finally, the efficiency of online voter registration depends upon the implementation of existing elections administration infrastructure in the state--a statewide electronic voter registration database in particular. Even though online voter registration disposes of manual data entry to get registrations into the statewide database, incompatible data-exchange formats may still require costly extra data-entry steps on the part of elections officials. (46)
C. Online Voter Registration Enhances Election Integrity
While the threat posed by voter registration fraud to election outcomes is all but nonexistent--researchers at the Brennan Center for Justice could not find one recent substantiated case in which a fraudulent registration led to a fraudulent vote cast (47)--it is certainly true that even isolated incidences of fraudulently submitted voter registration forms undermine election integrity. Fraudulently submitted forms force elections administrators to expend resources on investigation, give rise to costly litigation, and undermine public confidence in the system. Though the functional benefits of online voter registration may be more pronounced in increasing access and decreasing costs, the system's salutary effect upon election integrity gives it additional bipartisan appeal.
The safeguards against voter registration fraud in an online voter registration system are just as strong as under the existing paper card system--and in some respects, even more robust. Like paper form registrations, online registrations are verified against state identification cards and felony and death records, as well as against the registration records of neighboring Washington State. (48) But the online registration system's treatment of the signature requirement sets it apart from the current paper form system. Currently, a person attempting to submit a fraudulent paper voter registration application could provide a signature on the card that they could easily duplicate on the ballot envelope (should their fraudulent application somehow slip through the system). By contrast, under an online registration system, the registrant's state ID signature is used to verify the ballot envelope signature. A determined online fraudster must either fraudulently obtain a state ID from the DMV for every registration he plans to submit, or steal state IDs from eligible voters in order to copy their signatures--two substantially more difficult undertakings.
Online voter registration also addresses several other perceived risks of voter registration fraud. First, as noted above, online voter registration mitigates human errors associated with the processing of paper registrations and reduces "no-match" records, increasing public confidence in the system's integrity. Second, online voter registration reduces voters' reliance on third-party voter registration drives, which are vilified (usually with partisan overtones) for introducing large amounts of incomplete, illegible, or fraudulent registrations into the system. (49) While evidence of actual election fraud resulting from voter registration drives is virtually nonexistent, (50) "cracking down" on voter registration drives with online voter registration can provide political cover to policymakers with constituencies otherwise suspicious of the concept. (51)
The few election integrity problems specifically raised by online voter registration have more to do with the existing election administration system, rather than anything endemic to online voter registration itself. Indeed, most of the legislators who voted against HB 2386 indicated that they supported online voter registration but opposed Oregon's voter registration standards with respect to proof-of-citizenship requirements. (52) Moreover, the risks to election integrity posed by online attacks substantially overlap those faced by statewide voter registration databases. (53) To this end, the continued development of security standards and procedures for statewide voter registration databases will also improve the integrity of online voter registration systems.
While many states will likely jump on the online voter registration bandwagon in the coming years, system start-up costs pose the biggest obstacle to broader implementation. Congress can nudge the process along by passing House Bill 1719, the Voter Modernization Act of 2009, directing states to adopt online voter registration systems and allocating HAVA funds to enable them to do so. (54) Bipartisan passage of the bill would represent a fitting conclusion to a successful partnership between Democrats, Republicans, the federal government, and the states towards an election administration triple bottom line.
(1.) See Daniel P. Tokaji, Voter Registration and Institutional Reform: Lessons from a Historic Election, 3 HARV. L. & POL'Y REV. ONLINE, Jan. 22, 2009, at 1, http://www.hlpronline.com/Tokaji_HLPR_012209.pdf, at 1.
(2.) U.S. ELECTION ASSISTANCE COMM'N, THE IMPACT OF THE NATIONAL VOTER REGISTRATION ACT OF 1993 ON THE ADMINISTRATION OF ELECTIONS FOR FEDERAL OFFICE 2007-2008, at 1 (2009), available at http://www.eac.gov/program-areas/ research-resources-and-reports/completed-research-and-reports/program-areas/ research-resources-and--reports/copy_of_docs/ nvra-report-final_congress.pdf. The total number of registered voters jumped by 5.4% from 2004 to 2008, an increase of nearly ten million registrants, accompanied by millions more voters submitting voter registration applications to update their address or change their party affiliation. See Michael McDonald, George Mason Univ. U.S. Elections Project, 2008 General Election Voter Registration Statistics, http://elections.gmu.edu/Registration_2008G.html (last visited Jan. 30, 2010). A significant proportion, if not an outright majority, of the applications were received in the last few months before the voter registration deadline, leaving county elections officials scrambling to process the forms in time for the election. See, e.g., R. Michael Alvarez & Thad E. Hall, Resolving Voter Registration Problems: Making Registration Easier, Less Costly and More Accurate 9-10 (Caltech/MIT Voting Tech. Project, Working Paper No. 87, 2009), available at http://vote.caltech.edu/drupal/files/working_paper/ wp_87_pdf_4acfa68b61.pdf (showing that more Franklin County, Ohio voters registered between August and October 2008 than in the year before the March primary elections).
(3.) See generally U.S. Student Ass'n Found. v. Land, 546 F.3d 373 (6th Cir. 2008) (denying request to stay preliminary injunction against Michigan Secretary of State and Director of Elections from rejecting a voter's registration if the voter's identification card is returned as undeliverable); Common Cause of Colo. v. Buescher, No. 08-CV-02321-JLK, 2009 WL 1847353 (D. Colo. June 26, 2009) (alleging that Colorado officials violated NVRA by removing voters from statewide voter registration database); Am. Ass'n of People with Disabilities v. Herrera, 580 F. Supp. 2d 1195 (D.N.M. 2008) (denying preliminary injunction against enforcement of New Mexico's laws regulating 3rd party voter registration drives); Ass'n of Cmty. Orgs. for Reform Now v. Scott, No. 08-CV-4084-NKL, 2008 WL 2787931 (W.D. Mo. Jul. 15, 2008) (granting and denying injunctions against Missouri for alleged failure to distribute voter registration forms at public assistance offices).
(4.) See R. Michael Alvarez, Voter Registration: Past, Present, and Future 3 (Caltech/MIT Voting Tech. Project, Working Paper No. 30, 2005), available at http://vote.caltech.edu/drupal/files/working_paper/vtp_wp30.pdf (noting that typographical errors are inevitable in databases containing information about millions of voters).
(5.) I borrow the term "triple bottom line" from the corporate social responsibility expert John Elkington, who used it to describe an expanded set of values and criteria for measuring success--in the business context, criteria based not solely on the traditional profit bottom line, but also on social and environmental value added. See generally JOHN ELKINGTON, CANNIBALS WITH FORKS: THE TRIPLE BOTTOM LINE OF 21ST CENTURY BUSINESS (New Society 1998) (1997).
(6.) For a continuously updated list of states that have adopted or introduced online voter registration legislation, see Brennan Center for Justice, Voter Registration Modernization in the States: Online Registration, http://www.brennancenter.org/content/pages/vrm_online_registration (last visited Jan. 30, 2010).
(7.) See, e.g., Robert Bauer & Trevor Potter, Op-Ed., A New Page for Voting: It's Time to Ditch Paper-Based Registration, WASH. POST, June 25, 2009, at A19 (former general counsels to the Obama and McCain 2008 presidential campaigns) (arguing for the adoption of a paperless voter registration system); Ken Blackwell, Op-Ed., Time for a Big Overhaul: Reform Could Help Rid System of Manipulators, WASH. TIMES, Oct. 9, 2009 (former Republican Secretary of State of Ohio) (arguing that automated and online voter registration would "get rid of ACORN and ensure that only eligible citizens get on the rolls"); John Lindback, Op-Ed., Voter Registration: Time to Trade in Our Costly Clunker, OREGONIAN, Dec. 29, 2009 (senior officer for election initiatives at the Pew Center for the States, former Director of Elections for State of Oregon) (praising online voter registration as an innovative component of a modernized registration system).
(8.) Although some academics and policy experts have recommend adoption of online voter registration, few have conducted an in-depth analysis of the policy's strengths and weaknesses. See, e.g., PEW CTR. ON THE STATES, BRINGING ELECTIONS INTO THE 21ST CENTURY: VOTER REGISTRATION MODERNIZATION 3 (2009), available at http://www.pewcenteronthestates.org/uploadedFiles/ Voter_Registration_Modernization_Brief_web.pdf; Robert Richie, Leave No Voter Behind: Seeking 100 Percent Voter Registration and Effective Civic Education, NAT'L CIVIC REV., Fall 2007, at 39, 43. But see Modernizing the Election Registration Process: Hearing before the Subcomm. on Elections of the H. Comm. on H. Administrations, 111th Cong. (2009) [hereinafter Hearings] (statement of Wendy Weiser & Nicole Austin-Hillery, Brennan Ctr. for Justice), available at http://www.brennancenter.org/page/-/Democracy/Weiser Austin-Hillery H.R.1719 Testimony.pdf.
(9.) Although the Arizona legislature and Secretary of State deserve much credit as the first to successfully create an online voter registration system, they did so before the election administration controversies that arose out of the 2004 and 2008 elections, and before they or any other state had created a statewide online voter registration database
as required by the Help America Vote Act. I chose to focus on Oregon in the hope that it is more reflective of the political and policy concerns of similarly situated state legislatures in 2010 and beyond.
(10.) Mandated by section 303 of the 2002 federal Help America Vote Act (HAVA). See Help America Vote Act [section] 303, 42 U.S.C. [section] 15483 (2006).
(11.) However, turnout gains attributed to vote-by-mail appear to result from retention of existing voters in the system, rather than the recruitment of new voters. Cost savings of vote-by-mail are apparent in comparison to a hybrid system of absentee and polling place balloting. See PAUL GRONKE, CTR. FOR DEMOCRACY & ELECTION MGMT. AT AM. UNIV., BALLOT INTEGRITY AND VOTING BY MAIL: THE OREGON EXPERIENCE 3-4 (2005).
(12.) In 2004, the turnout gap between 18- to 24-year-old Oregonian voters and voters over 25 was 28%, one of the largest gaps in the nation. See CARRIE DONOVAN ET AL., CTR. FOR INFO. & RESEARCH ON CIVIC LEARNING & ENGAGEMENT, YOUTH VOTER TURNOUT IN THE STATES DURING THE 2004 PRESIDENTIAL AND 2002 MIDTERM ELECTIONS 3 (2005), available at http://www.civicyouth.org/PopUps/FactSheets/FS_04_state_vote.pdf.
(13.) Ben Cannon, Or. State Legislature House of Representatives, Floor Speech in Support of HB 2386 (March 20, 2009) (notes on file with author). Before 2010, Oregonian registrants could fill out a paper voter registration form to send to or drop off at a county elections office or the Secretary of State's office, or register at the Department of Motor Vehicles and other public agencies as provided under the federal National Voter Registration Act (NVRA), commonly known as "Motor Voter." See Oregon Secretary of State, Voting in Oregon Guide: Registering to Vote, http://www.uhavavote.org/ votingguide/register.html (last visited Feb. 19, 2010).
(14.) The bill, which died in committee, would require first time registrants to supply documentary proof of citizenship to complete a voter registration. See H.B. 3432, 75th Leg. Assem. (Or. 2009).
(15.) H.B. 2386 [section] 2(1)(a)-(c), 75th Leg. Assem. (Or. 2009).
(16.) Id. [section] 2(4).
(17.) STAFF OF OR. LEG. FISCAL OFFICE, 75 LEG. ASSEM., FISCAL ANALYSIS OF H.B. 2386 B-ENOROSSED (2009). This estimate falls in the middle of start up costs projected by other states, ranging from $100,000 in Arizona to $400,000 in California. See Hearings, supra note 8, at 4.
(18.) For a discussion of "No" votes, see infra note 52.
(19.) The proposed rules includes a provision for an electronic time and date record at the time of submission that will be deemed the time submitted for voter registration purposes, and for the county elections offices to process and accept or reject the registration application submitted online. See Notice of Proposed Rulemaking: Adoption of Online Voter Registration System (proposed Jan. 8, 2010) (to be codified at OR. ADMIN. R. 165-005-0160).
(20.) See Michael McDonald, George Mason University U.S. Elections Project, 2008 General Election Voter Registration Statistics, http://elections.gmu.edu/Registration_2008G.html (last visited Jan. 30, 2010).
(21.) Although there remains a lively academic debate about the precise effects of different voter registration laws and policies, there is broad consensus that registration requirements and procedures generally decrease participation among otherwise eligible voters. See, e.g., Daniel P. Tokaji, Voter Registration and Election Reform, 17 WM. & MARY BILL RTS. J. 453, 496 (2008); Stephen Ansolabehere & David M. Konisky, The Introduction of Voter Registration and its Effect on Turnout 16 (Caltech/MIT Voting Tech. Project, Working Paper No. 14, 2004).
(22.) R. MICHAEL ALVAREZ ET AL., PEW CTR. ON THE STATES, 2008 SURVEY OF THE PERFORMANCE OF AMERICAN ELECTIONS 34, 59 (2009), available at http://www.pewtrusts.org/uploadedFiles/wwwpewtrustsorg/Reports/ Election_reform/Final% 2520report20090218.pdf.
(23.) See JENNIFER ROSENBERG & MARGARET CHEN, BRENNAN CTR. FOR JUSTICE, EXPANDING DEMOCRACY: VOTER REGISTRATION AROUND THE WORLD 2 (2009), available at http://www.brennancenter.org/page/-/publications/Expanding.Democracy.pdf (citing laws in fourteen other industrialized countries that either required citizens to register to vote or saw the government take an active role in keeping voter rolls up to date).
(24.) For example, in one nonpartisan youth voter registration drive in Oregon in 2008, almost a third of voter registration applications collected came from already-registered Oregonians who simply needed to update their address. The share of in-state movers registered by the Oregon Bus Project was a higher percentage of total registrants than any other comparable nonpartisan field program in the nation. See ETHAN ROEDER, NEW ORGANIZING INSTITUTE, VOTER REGISTRATION ANALYSIS '08: EVALUATING INDEPENDENT VOTER REGISTRATION EFFORTS FROM THE 2008 ELECTION CYCLE 34 (2009).
(25.) Even though most Americans believe that registering to vote is not a difficult task, younger voters, voters of color, and voters with lower levels of education are more likely to view voter registration as difficult compared to older, white, and better-educated voters. See R. Michael Alvarez et al., How Hard Can It Be: Do Citizens Think it Is Difficult to Register to Vote? 18 STAN. L. & POL'Y REV. 382, 397-400 (2007). Another corollary factor is differential access of demographic groups to the Motor Voter system. While many Americans register to vote while obtaining or updating a driver's license at the DMV, younger citizens who obtain their driver's licenses two years before they are eligible to vote, lower income citizens who rely exclusively upon public transportation, and disabled and elderly citizens who can't drive are all less likely to take advantage of this convenient process.
(26.) A good number of Oregonians lack life-long experience with state specific voter registration practices; only about half of the population was born in the state, one of the lowest percentages in the nation. See R. Michael Alvarez et al., Interstate Voter Registration Database Matching: The Oregon-Washington 2008 Pilot Project 2 (Caltech/MIT Voting Tech. Project, Working Paper No. 84, 2009), available at http://www.vote.caltech.edu/drupal/files/working_paper/ wp_84_pdf_4acf7a043a.pdf.
(27.) Even with special federal and state laws that relax registration and voting deadlines for military service members, overseas and active duty service members register and vote at lower rates than the civilian population. See ADAM SKAGGS, BRENNAN CTR. FOR JUSTICE, REGISTERING MILITARY AND OVERSEAS CITIZENS TO VOTE 3 (2009), available at http://www.brennancenter.org/page//Democracy/Registering%20Military%20 and%20Overseas%20Citizens%20to%20Vote.FINAL.pdf.
(28.) See KARLO BARRIOS MARCELO, CTR. FOR INFO. & RESEARCH ON CIVIC LEARNING & ENGAGEMENT, FACT SHEET: VOTER REGISTRATION AMONG YOUNG PEOPLE 2 (2008), available at http://www.civicyouth.org/PopUps/FactSheets/ FS07_Registration.pdf.
Unsurprisingly, states that allow people to register to vote on election day report the highest levels of turnout in the nation for voters of all demographic backgrounds. See DEMOS, DEMOS POLICY BRIEF: VOTERS WIN WITH SAME DAY REGISTRATION 1, 5 (2009), available at http://www.demos.org/pubs/voterswin_09.pdf.
(29.) See James G. Gimpel et al., Election-Year Stimuli and the Timing of Voter Registration, 13 PARTY POL. 351, 356-59 (2007). What's more, people who register late vote at a higher rate than those who register earlier in the year--as high as 7% in one study. See Expert Declaration of Michael P. McDonald at 7, League of Women Voters of Fla. v. Browning, 575 F. Supp. 2d 1298 (S.D. Fla. 2008); Gimpel et al., supra, at 368. Enhancing the ability of late registrants to successfully submit their voter registration cards thus has a considerable "bang for your buck" in terms of increasing participation.
(30.) Indeed, online voter registration seems to be popular with young voters in particular. In the first six months of implementation, over half of Washington's online voter registrants were between the ages of eighteen and thirty. See OFFICE OF THE WASH. SEC'Y OF STATE, WASHINGTON STATE ONLINE VOTER REGISTRATION 6 (2009), available at http://nass.org/index2.php?option=com-docman&task=doc-view& gid=632&Itemid=99999999.
(31.) For example, more than forty-two million Americans between the ages of eighteen and thirty-four are active users of Facebook, and half of active Facebook users log in on any given day, for an average of fifty-five minutes. See Justin Smith, Latest Data on Facebook's U.S. Growth by Age and Gender, INSIDE FACEBOOK (Oct. 1, 2009), http://www.inside facebook.com/2009/10/01/latest-data-on-facebooks -us-growth-by-age-and-gender-october-1-2009/; Facebook Statistics, http://www.facebook.com/press/info.php?statistics (last visited Jan. 30, 2010).
(32.) What's more, the Facebook user receives the registration form from a trusted peer messenger, rather than from an anonymous government employee or an unfamiliar street canvasser.
(33.) See, e.g., Benjamin Highton, Easy Registration and Voter Turnout, 59 J. POE. 565, 573 (1997); KEI KAWASHIMA-GINSBERG ET AL., CTR FOR INFO. & RESEARCH ON CIVIC LEARNING & ENGAGEMENT, FACT SHEET: STATE ELECTION LAW REFORM AND YOUTH VOTER TURNOUT 2 (2009), available at http://www.civicyouth.org/PopUps/FactSheets/State_law_and_vouth_ turnout_Final.pdf.
(34.) See Alvarez & Hall, supra note 2, at 4 (noting debate in the research literature about whether voter registration reforms result in additional voter participation, or have just made it easier for already-high-propensity voters to participate); Pamela S. Karlan & Eban Moglen, The Soul of a New Political Machine: The Online, the Color Line and Electronic Democracy, 34 LOY. L.A.L. REV. 1089, 1089 (2001).
(35.) No empirically rigorous study has yet measured the independent turnout effect of online voter registration, but even an optimistic estimate might only approach the 4.7 to 8.7 percent increase attributed to Motor Voter laws. See Benjamin Highton & Raymond E. Wolfinger, Estimating the Effects of the National Voter Registration Act of 1993, 20 POL. BEHAV. 79, 79 (1998) (discussing estimated independent turnout effect of NVRA implementation).
(36.) The most obvious example of litigation resulting from problems with voting machines is Bush v. Gore, 531 U.S. 98 (2000).
(37.) Even in the much criticized 2000 election, where faulty voting equipment and confusing ballots accounted for 1.5 to 2 million votes lost, voter registration problems accounted for a larger share of lost votes--l.5 to 3 million. See CALTECH/MIT VOTING TECH. PROJECT, VOTING: WHAT IS, WHAT COULD BE 9 (2001), available at http://www.vote.caltech.edu/drupal/files/report/voting_what_is_what _could_be.pdf. Thanks to federal and state elections administration reforms (such as HAVA), the impact of "voting booth problems" on lost votes has been significantly reduced--the Caltech/MIT Voting Technology Project report for the 2008 election did not even list voting equipment and confusing ballots as a source of lost votes. But the number of lost votes attributable to registration problems was almost as high in 2008 as it was in 2000. See Alvarez & Hall, supra note 2, at 3.
(38.) See R. MICHAEL ALVAREZ ET AL., supra note 22, at 58-59.
(39.) One virtue of vote-by-mail is that county elections divisions are in more frequent contact with voters, leading to cleaner and more accurate voter registration records.
(40.) As many as four to six percent of ballots are undeliverable in Multnomah County, the most populous county in Oregon. See GRONKE, supra note 11, at 5.
(41.) A 2004 audit of the New York City elections division yielded an almost twenty percent error rate in the transcription of drivers' license numbers alone. See JUSTIN LEVITT ET AL., BRENNAN CTR. FOR JUSTICE, MAKING THE LIST: DATABASE MATCHING AND VERIFICATION PROCESSES FOR VOTER REGISTRATION 23 (2006), available at http://www.brennancenter.org/dynamic/subpages/download_file_49479.pdf.
(42.) Arizona experienced just such a problem before their primary election voter registration deadline in 2008. See Paul Davenport, Online System Malfunctions in Advance of AZ Registration Deadline, ASSOCIATED PRESS, Jan. 8, 2008. However, the adoption of online voter registration does not foreclose existing methods of voter registration. Moreover, it is doubtful that even a major computer failure near a voter registration deadline would result in a lower level of voter registration than in the absence of online voter registration altogether; the subset of citizens who would have registered online but would not seek out a paper registration form in the event of a computer crash are probably unlikely to register in the absence of an online option--with overseas, out of state, and very-last-minute registrants being the troubling exceptions.
(43.) See PEW CTR. ON THE STATES, THE REAL COST OF VOTER REGISTRATION: AN OREGON CASE STUDY 5 (2009), available at http://www.pewcenteronthestates.org/uploadedFiles/The_Real_Cost of Voter_Registration.pdf.
(44.) This figure likely underestimates the total savings per transaction, since it does not appear to include the costs per application borne by the Secretary of State--printing and distributing voter registration cards chief among them. Since over seventy percent of Arizonans register to vote online, the cost savings are quite substantial. See PEW CTR. ON THE STATES, supra note 8, at 3.
(45.) Author's estimate based on Oregon's 1,152,761 voter registration transactions in 2008. See PEW CTR. ON THE STATES, supra note 43, at 5.
(46.) See R. MICHAEL ALVAREZ & THAD HALL, IBM CTR. FOR THE BUS. OF GOV'T, THE NEXT BIG ELECTION CHALLENGE: DEVELOPING ELECTRONIC DATA TRANSACTIONS STANDARDS FOR ELECTION ADMINISTRATION 19 (2005), available at http://www.businessofgovernment.org/pdfs/AlvarezReport.pdf. For example, the statewide voter registration database and the state agency that maintains death records may have different formats for date of birth (mm-dd-yyyy versus dd-mm-yy) or for dealing with multiple-word last names (Elizabeth De La Torre versus Elizabeth Delatorre). Though section 303 of HAVA requires that the database "be coordinated with other agency databases within the State," the statute does not mandate that state elections agencies coordinate data-exchange formats with state agencies or other states' elections divisions, nor does it provide specific funding to do so. Help America Vote Act [section] 303, 42 U.S.C. [section] 15483(a)(1)(A)(iv) (2006). Unsurprisingly, observers note a lack of interest and incentive on the part of various state agencies to collaborate on creating database interoperability with the state's elections division. See COMM. ON STATE VOTER REGISTRATION DATABASES, NAT'L RESEARCH COUNCIL, IMPROVING STATE VOTER REGISTRATION DATABASES FINAL REPORT 17 (2009), available at http://www.nap.edu/catalog/12788.html. The costs of correctly formatting and matching information are thus borne by the elections division, the agency that bears the ultimate statutory responsibility to ensure that applicants attempting to register are correctly verified. For example, in Oregon the Secretary of State sent its HAVA funds to the DMV to set up the transfer of signature files for online voter registration. County elections divisions also must bear the costs of attempting to contact non-matched voters to verify their eligibility.
(47.) See JUSTIN LEVITT, BRENNAN CTR. FOR JUSTICE, THE TRUTH ABOUT VOTER FRAUD 20 (2007), available at http://truthaboutfraud.org/pdf/TruthAboutVoterFraud.pdf.
(48.) See generally Alvarez et al., supra note 26 (noting that Oregon and Washington are two of a handful of states that are cross-referencing their statewide voter registration databases against each other to identify voters registered to vote in both states).
(49.) See, e.g., Blackwell, supra note 7.
(50.) See LEVITT, supra note 47, at 18.
(51.) See, e.g., Hearings, supra note 8, at 4, available at http://cha.house.gov/UserFiles/245_testimony.pdf (statement of Todd Rokita, Indiana Secretary of State) (criticizing draft of federal online voter registration legislation that, unlike Indiana's system, does not require the use of DMV signature files, on the grounds that such a system would be vulnerable to fraud).
(52.) See, e.g., Jim Thompson, Or. State Legislature House of Representatives, Vote Explanation on H.B. 2386, Mar. 26, 2009 (on file with author). Minority Leader Bruce Hanna and Representatives Vic Gilliam, Andy Olson, and Tim Freeman submitted identical letters to the Chief Clerk of the House of Representatives. Almost every "No" vote against HB 2386 can be attributed to the concern on the part of a group of Representatives that non-citizens who had acquired a state ID under since-repealed DMV regulations could register to vote online. Before Oregon adopted a law requiring proof of citizenship to obtain a state ID in 2006, non-citizens could theoretically obtain drivers' licenses, and using those drivers' licenses, register to vote. Evidence of such action is not forthcoming, and the theoretical window in which such a felony could occur is fast closing; subsequent state ID re-registrations will require proof of citizenship as well. As advocates pointed out, non-citizens foolish enough to submit identifying information to the government and risk felony charges and deportation in exchange for casting one marginal vote could do so under the current paper card system.
(53.) Such risks include exploitation of the data transmission process between state databases to register fictitious voters, delete or alter registration records, or disrupt the transfer of registration information from the Secretary of State to county elections offices. See generally R. Michael Alvarez, Potential Threats to Statewide Voter Registration Systems (Caltech/MIT Voting Tech. Project, Working Paper No. 40, 2005), available at http://www.vote.caltech.edu/drupal/files/working_paper/vtp_wp40.pdf. As noted above, an online voter registration system could be uniquely vulnerable to a denial of service attack by someone seeking to prevent registrants from submitting their registrations online, a risk that could be mitigated by fail-safe backup servers.
(54.) See Voter Registration Modernization Act of 2009, H.R. 1719, 111th Cong. (2009).
Alex G. Tischenko, B.A. 2007, Pomona College; J.D. Candidate 2012, Stanford Law School; former Legislative Aide to Oregon State Representative Ben Cannon and Lead Field Organizer in Oregon Bus Project's 2008 youth voter registration drive.