Matthew Waring and David Waterman, Interns, AEI-Brookings Election Reform Project
September 3, 2008
As Election Day 2008 approaches, the nation's focus will be on the two candidates with their last-minute appeals to the voters, and on the pundits with their predictions and electoral maps. What will go largely ignored will be the complex infrastructure that is needed for the election to take place, and the massive volunteer force of poll workers needed to ensure it runs smoothly. These workers usually train for only a few short hours and serve just one day, but their impact on the election can be enormous. They are entrusted with numerous responsibilities, from signing in registered voters and explaining voting procedures to operating voting machines (sometimes multiple types in the same precinct) and handling ballots. As one former Virginia election official aptly put it, "On Election Day, the poll worker is God." Consequently, having inadequately trained staff at the polling place can lead to inefficiency and long lines, or worse, the unintentional disenfranchisement of voters because of errors. Thorough training is also essential because poll workers can play a role in shoring up public faith in the workings of the political system; A study by Thad Hall, J. Quin Monson, and Kelly Patterson, for example, found that a voter's assessment of the quality of her poll workers is a strong predictor of her confidence in the integrity and fairness of the electoral process.
[1]
Since the controversial elections of 2000 and 2004, efforts to make American elections more fair and transparent have largely focused on improving the technology involved, with particular concern of late placed on the development of systems that produce a record of each vote, but poll worker issues are beginning to receive the attention they deserve. In August 2007, the U.S. Election Assistance Commission (EAC) issued a
guide to "Successful Practices for Poll Worker Recruitment, Training and Retention," and EAC chair Rosemary Rodriguez has
called this year for the recruitment of two million poll workers in anticipation of a record turnout this November.
One obstacle to the study and improvement of poll worker training is the diversity and, in some ways, opacity of the preparation process. The EAC and state governments can and do issue useful general guidelines, but local administrators-for whom election administration is often only a small part of the job--are the ones who actually devise and execute training programs. The result is an incredible diversity of practices, each with, presumably, varying degrees of effectiveness. In order to get a sense of some of the diverse practices that exist, we examined how a number of jurisdictions within Virginia, Maryland, and the District of Columbia deal with poll workers with the hope of getting a better sense of how jurisdictions on both sides of the nation's capital approach the recruitment and training of poll workers. We ultimately found practices both worthy of emulation, and ones that need improvement.
The ten jurisdictions we examined in Virginia and Maryland displayed considerable variation both in poll worker training methods and in their practices for recruiting and compensating those workers. Broadly speaking, the practices of each jurisdiction within a given state are similar because each must follow the same state election law, and each begins with the same set of core materials issued by the state electoral board. Inevitably, however, each county or municipality makes modifications to that common foundation in response to its own needs.
It seems intuitive that compensation and incentives should be a critical part of recruiting and retaining an adequate number of poll workers. The majority of counties and municipalities in Virginia and Maryland pay workers a small stipend for attending one or more mandatory training sessions in addition to their pay for working Election Day, but this practice is not universal. The wage rate for Election Day itself also varies from locality to locality. In Fairfax County, VA, for example, poll workers are not paid for being trained and thus compensated only for their work on Election Day. Pay in Fairfax is $100 for election officers and $200 for precinct chiefs. Across the Potomac River in Montgomery County, MD, workers are paid for both training sessions and for Election Day; pay for chief judges is $200, with an additional $50 for training, and $150 for regular election judges, with $30 for training.
The city of Charlottesville, VA has expanded the scope of poll worker compensation even further. For over a decade, the city has purchased bag lunches to distribute to its poll workers on Election Day. As a part of this initiative, the Voter Registration and Elections Office in Charlottesville has formed a cooperative relationship with a local downtown area vendor and restaurant. The city's outreach coordinator, Jackie Harris, noted that there is a high retention rate for poll workers in Charlottesville, and with many poll workers working a 14-hour day with little chance to leave the polling place, it may be that providing free food can turn an arduous experience for workers into a more pleasant one. This service not only saves workers the trouble of bagging their own lunches (as they must do in many other localities), but also is a way that the city can show its gratitude for their efforts.
Diane Gilliland, the deputy registrar in Charlottesville, put forth an alternative explanation for the city's high retention poll worker retention rate, claiming that the explanation involves the "great camaraderie in each precinct; the poll workers, who come from diverse backgrounds and are of different ages, seem to really like each other." According to Ms. Gilliland, a few friendly, older women and one popular police officer, who volunteer as election judges, bake cookies on Election Day to motivate the poll workers to put forth their best effort, which never fails to "warm everyone's hearts." In general, smaller jurisdictions have an easier time encouraging this type of atmosphere; in Allegany County, MD, for example, individual precinct teams often attend training as a unit to practice with voting machines. This sort of collegiality can only improve both the attitude and the effectiveness of poll workers. Obviously, larger localities cannot always provide exactly the same feel, but having precinct teams train together when possible would be a useful way to simulate it.
In terms of recruitment, most counties have supplemented their own efforts by developing relationships with community organizations and the corporate world. In the City of Charlottesville, for instance, the election office has sought out corporate sponsorship with the goal of recruiting 10,000 officers through its "Ensure the Integrity of the Vote" initiative. Among the city's partners is State Farm Insurance, which has contributed to this effort by granting administrative leave for those employees who wish to serve as election judges, a technique also used with great success by many of the other local governments we examined.
Allegany County, MD, has employed a different technique, developing a tradition of recruiting poll workers through local organizations, like the Rotary Club; recruiting in this way has the potential to be particularly effective, since members of such organizations are likely to be civic-minded. The county also has recruiting operations at local universities, including Allegany College of Maryland and Frostburg State University. In recent years, college students at Allegany College helped with voter and absentee voter registration. Undergraduates are a particularly appealing body of potential poll workers for local election officials, since young adults tend to be more experienced with technology and thus better able to work with newer, perhaps computerized, voting systems. To its credit, the EAC has picked up on the fact that young people are often especially tech-savvy, and in June 2008
gave out $750,000 in grants to colleges in 18 states for efforts to recruit more students as poll workers.
Despite these successes, there are still some challenges to overcome in recruiting students; many are normally stuck in class or otherwise occupied by studies and thus are unable to work long hours on Election Day. Split shifts, which unfortunately have yet to be offered to poll workers in many of the jurisdictions we surveyed, may provide an opportunity for these students and others with similarly demanding jobs or commitments. Such policies allow poll workers to work half days after finding an alternate at their training session to work either the morning or afternoon, thus avoiding a full day of service if they cannot spare it.
Various jurisdictions also actively recruit poll workers from the general public in many other ways: public service announcements, direct mail, presentations at high schools, through party central committees, referrals by word-of-mouth, media advertising, the Internet, and through a system of referrals. Montgomery County, MD has even begun recruiting at the Sugar Loaf Mountain Wine Festival, which is known to draw a large number of registered voters. In terms of recruiting, therefore, one clear trend is that the more numerous and varied the methods, the better the results; thus, efforts by states to share innovative recruiting "best practices" statewide would be useful sources of ideas for jurisdictions needing to ramp up their recruiting efforts.
Our findings also suggest that rural localities, such as Allegany County, have a much easier time recruiting and retaining their ideal number of poll workers whereas in more densely populated areas, such as Loudoun County, Virginia, the electoral boards have a much more difficult time. According to Jeff Mangold, an administrator at the Loudoun County electoral board, the county, a suburb of Washington, DC, "desperately" needs a couple hundred more poll workers for the upcoming presidential election. Rural counties tend to have much less of a problem, perhaps because they simply need fewer poll workers . Another difficulty, moreover, arises when jurisdictions, because of state law, are required to have poll workers of different partisan affiliations. For example, in Montgomery County, which is highly Democratic, officials have had a difficult time locating willing Republicans to work as election judges.
Not surprisingly, all jurisdictions surveyed indicated that they would be increasing the total number of poll workers or election judges for the 2008 presidential election, for which there is expected to be record turnout. Prince George's County, Maryland explained that generally for each voting cycle they require 2,800 to 3000 election judges. Based on the interest in the upcoming election, however, the county will be recruiting an additional 300 to 500 judges. Likewise, Arlington County, Virginia generally employs somewhere between 500 to 600 poll-workers in non-presidential election years; for the upcoming presidential election, the office expects to hire around 850 to 1,000 poll-workers.
Aside from concerns about how best to attract and retain a critical mass of poll workers, the other important issue in poll worker administration is the question of training methods. When it comes to training, the growing consensus seems to be that more innovative and interactive methods produce better results. In their 2007 study published in
PS: Political Science and Politics, Hall, Monson and Patterson compared the results of training in Cuyahoga County, Ohio, which uses traditional lecture-style methods in classes of 30 to 40, with training in Utah, which provides more hands-on experience with equipment in classes of 14 to 16. They found that Utah poll workers generally felt more prepared for Election Day and were less likely to notice a difference between the training and how the equipment functioned at the actual polling place.
[2]
Unfortunately, in the jurisdictions we examined, such interactive methods have only begun to make serious headway. Most still rely primarily on a lecture format, supplemented by a lengthy manual, which, one suspects, only the most dedicated will read all the way through. Almost all incorporate a voting machine and electronic poll book into their training for demonstration purposes, but few have made the sort of commitment to hands-on training that we found in Montgomery County, MD, which sets aside 80 voting machines for use in training, so that each trainee can practice with one even when multiple classes are occurring simultaneously. Most do make voting machines available at the election office on a walk-in basis for trainees to get additional practice, but only if the poll workers feel they need it. Some of the larger jurisdictions would probably benefit from implementing something similar to the EAC's "Practice Makes Perfect" program; employed in Utah, the program ensures multiple practice stations are scattered throughout the jurisdiction in locations such as libraries. Lowering the cost of travel to a practice location would encourage more workers to take advantage of the opportunity.
Another innovation that still appears to be a long way off is online training. Online components can offer much that written manuals and guides cannot, including videos and animations of poll worker tasks, as well as interactive quizzes that enable trainees to receive instant feedback (and allow election officials to gauge which topics are not being covered effectively in training). The state of Texas has been a leader in this area; last year, it rolled out an
extensive online system complete with case studies, tests, and even "Certificates of Completion" for trainees to print out. Unfortunately, hardly any of the jurisdictions we contacted currently use online materials in the training process. The major barrier to the introduction of online materials, however, seems to be cost, rather than rigidity or apathy on the part of local election officials. Jeff Mangold, an administrator in Loudoun County, VA, said that Loudoun "would love" to have such material available, but the resources and staff necessary to implement such a project simply are not available. Many other officials expressed similar sentiments. If funding for creating and maintaining online training tools ever becomes available, we will likely see them deployed much more extensively.
The successful administration of elections is a process that continues to present real challenges to governments nationwide, and these challenges are likely to become even more daunting as turnout for federal elections swells. The Help America Vote Act of 2002 (HAVA) was an admirable first step in offering assistance to local governments trying to make elections run smoother and more fairly, but so far, the human part of our election machinery has received relatively little attention. A renewed commitment from all levels of government to this "human element" - in the form of greater resources devoted to training, recruitment, and compensation, and the use of smarter methods for training and evaluation - might revolutionize our poll worker force and bring us that much closer to the goal of an error-free election.
Matthew Waring and David Waterman were summer 2008 interns with the AEI-Brookings Election Reform Project. Comments about this article can be sent to politicalcorner@aei.org.
Viewpoint is an occasional feature analyzing various election reform issues.
Special thanks to election officers from the following jurisdictions: In Maryland, Allegany County, Baltimore City, Dorchester County, Harford County, Montgomery County, Prince George's County, and St. Mary's County, and in Virginia, Arlington County, Charlottesville City, Giles County, and Loudoun County