Election Delays

The data presented in this section come from two sources: the Survey of the Performance of American Elections (SPAE); and the Cooperative Congressional Election Study (CCES). Full reports based on this data will be released in the near future. For the purposes of this report, we extract several highlights below.

The first notable data finding is that based on a broad comparison with the 2008 election, voters in 2012 did not wait significantly longer at the polls on average. The percent of people waiting more than 30 minutes decreased compared to 2008. However, officials and journalists, cited throughout this document, reported massive delays scattered across the country in 2012. They impacted a significant number of voters. The data aggregating the total wait times fails to capture this situation. The longest waits occurred in Florida (45 minutes), the District of Columbia (33.8) Maryland (28.8), South Carolina (24.8), Virginia (23.6), Michigan (21.9), and Louisiana (20.4). Washington and Oregon hold all-mail elections and therefore have zero wait times.

Again, it is important to highlight that even within the states that experienced delays, some jurisdictions featured significantly longer waits than others. For example, in Florida certain counties had average waits upwards of 80 minutes, while neighboring counties had waits of only 7-10 minutes. Florida's average wait time of 45 minutes in 2012 glosses over the significantly higher wait times experienced in several areas. Notably, Miami, Lee County, and Orange County experienced over 80 minute wait times. Lee County arguably experienced the largest administrative meltdown of any county in Florida. There is a correlation between population density and wait times; as the population density increases, wait times also increase. This could account for variation within states where large cities may have experienced delays, whereas rural and suburban areas did not. However, population density is not a determinant factor as many urban centers experienced low wait times. For example, Los Angeles County averaged 7.7 minute waits on Election Day 2012. It is also important to note that areas with early voting also found increased wait times in more densely populated areas. Examining individual voters' race also seems to correlate with wait times. M.I.T. Political Science Professor Charles Stewart has observed that Black and Hispanic voters faced significantly longer wait times. According to Professor Stewart, in statistical analyses of the difference between White and Black voting wait times that control for where voters live with greater specificity (e.g., entire state versus county, versus to ZIP code), the gaps between White and Black voters shrink. As Professor Stewart explains, "White voters who live in areas with black voters wait longer to vote than white voters who live around other white voters." This suggests that while race may be a factor, it is not the only factor at play. While this Report does not specifically discuss the relationship between race and delays, some commentators have suggested the issue should be explored further. An analysis of this data is beyond the scope of this Report. Comparing individual wait time by partisan affiliation also yielded interesting results. There appear to be divides between Democrat and Republican average wait times; strong Democrats had the longest wait time (15.6 minutes) and strong Republicans the shortest (11.4 minutes). Like race, partisan correlations to wait times for voting are tied to multiple factors. Also like race, partisan impact of wait times at polling places has captured media interest.

Another subject of study was when within the Election Day administrative process delays occurred. It appears that the majority of delays occurred before voters checked in to vote. This data suggests that delays in 2012 were most often the result of problems that impede efficient voter check in such as voters attempting to vote at the wrong polling place, ID confusion, poll book inefficiencies, and voter registration errors. Professor Stewart indicated that these findings could challenge scholars' and policy analysts' previous conclusions that the number of voting machines at precincts is the most important factor causing delays. If the problem were voting machine shortages, the longest lines would presumably form after the check-in station. The SPAE results demonstrate, however, that the longest wait times occur before or during check-in. This could be caused by confusion regarding new voter ID laws. Another potential cause of delays identified was the decision of some jurisdictions not to reassign voters to new precincts following the post-census redistricting. For example, in Miami-Dade County, Florida, the local supervisor of elections decided to use split precincts to avoid voter confusion over newly assigned precincts. This combination required that additional steps be taken during the check-in process to ensure that each voter received the appropriate ballot style, causing further delays during check in. When translating concerns about precinct efficiency into policy recommendations, the SPAE data suggests that more focus should be directed to inefficiencies in check-in procedures and check-in technologies.

Overall, data from the 2012 election relating to wait times requires further study and analysis. Still, these preliminary findings are helpful in orienting the discussion that follows.

Contributors to Delay

The following sections provide an overview of the main contributors to election delays in 2012. This is by no means an exhaustive list, and indeed many of the factors discussed are intertwined.

One aspect of delays not singled out in the discussion below is election budgets--an inescapable precursor to smooth election administration in this country. However, assessing budget allocations and shortfalls in U.S. elections is a near impossible task for the simple reason that states do not report data on the amount of money spent on elections. Part of the reason is that states in which election administration is decentralized might not compile this information. Other reasons relate to the historic lack of data about elections in this country generally. As state election administrators begin to improve transparency of budgets, budget allocations and spending for elections, a clearer picture will emerge about the impact of spending on election delays.

Event Management
Elections are not different from any other event that attempts to engage a large number of people in a communal process. Events like weddings, concerts, and professional sports games require careful planning to ensure that the event runs smoothly. Arguably, elections require even more planning because of their vast scale and short timeframe.

From media reporting on the 2012 elections, it appears that many of the instances of delay can be attributed to problems in event management. Reports indicate that before the election, many precincts lacked sufficiently accurate estimates of the number of voters expected on Election Day. These estimation problems may have led to understaffing in some precincts. Some counties used vote centers instead of the precinct model, but did not anticipate that the campaigns and promote the vote organizations would direct voters to only one. Others planned inadequate spaces or numbers of locations for the activity of voting.

While some states have developed careful planning mechanisms for coping with weather and other emergencies (indeed many state election statutes require such planning), less drastic disruptions to voting also require adequate planning in advance. Election workers are often underprepared for mundane problems like machines breaking, power loss, or voters showing up at the wrong precinct or in high volume during particular periods of the day. The 2012 election provided examples of delays potentially related to poor planning. Even absent freak emergencies, voting machines will undoubtedly break, voters will arrive at the wrong precinct, and poll workers will arrive late. Problems always arise on Election Day.

Election managers' ability to predict turnout can hinge on how states define precincts. Some states, like Kansas, define precincts geographically regardless of population. Most states place a cap on the number of voters per precinct or booth. Capping the maximum number of people assigned to each precinct may ameliorate the problem of unexpected turnout. Although most states have per-precinct or per-booth voter cap laws, these laws will not prevent delays in elections with unprecedented voter turnout. Inadequate deployment of other resources (such as poll workers, poll books, and voting machines) is also a factor that voter cap statutes will not address.

How states address voter confusion is part of the event management process--do your guests know where to go and what to bring? These questions are explored in the Voter Confusion section below.

Summary of Proposed Solutions
Applying blanket event management solutions to all states and all precincts is ill-advised. Still, injecting management strategies borrowed from businesses adept at queuing customers may help election managers allocate resources to best address the needs of their jurisdiction. Thus far, commenters have suggested broad recommendations to enhance election administrators’ party planning.

1. Modernize voting/voter registration
2. Look to other industries 
3. Collect accurate election data
    

Voting Flexibility
For purposes of this section, voting flexibility relates to alternative methods of voting that states make available to electors aside from casting their ballot at the polls on Election Day. This section does not consider accommodations made for overseas and military voters, but rather focuses solely on whether and the extent to which states provide alternate methods of voting to voters residing in the state, such as "early voting" meaning in-person voting on days prior to the day of the general election; "no-excuse" early voting, in which a voter can come to a polling station prior to the election and cast a ballot without providing a reason for being absent on Election Day, "excuse-required" early voting, in which the voter must provide a reason for being unable to come to the polls on Election Day, and finally absentee voting, which refers to a citizen voting via mailed-in ballot prior to Election Day. The no-excuse/excuse dichotomy also applies to absentee voting.

While this Report discusses voting flexibility and notes some reforms' potential to reduce delays, it also recognizes that providing more voting options is not a solution without flaws. Introducing alternative methods of voting can cause confusion for voters who must keep more procedures straight. Similarly, it becomes harder for states to administer elections when there are more laws and regulations to track, and more technical information that must be passed on to poll workers about voting alternatives. Some of the measures of voting flexibility add to elections costs or risk security as well. The specific pros and cons of possible reforms are laid out below in the section entitled "Proposed Solutions."

Lastly, it is difficult to find direct evidence that certain states experienced Election Day delays because they did not allow alternate methods of voting, such as no-excuse absentee, early voting, or all-mail elections. Unlike a technology malfunction for which a voter can report experiencing a concrete problem, there are no anecdotes on the subject of flexibility because those methods of voting were simply not an option for these electors. Therefore, the best evidence of a lack of alternative voting options causing delays are inferences drawn from the correlation between wait times and the state legislation in place in 2012, and from comparisons with past years if there have been changes to the laws.

The relationship of delays to voter flexibility requires further study. Logic suggests that if the more people who are able to vote absentee or during a period of early voting, there will be fewer voters at the polls on Election Day. However, strained election budgets, shifting legislative landscapes, and inadequate voter education are likely to stymie improvements on Election Day wait times that increased voting flexibility might bring.

Summary of Proposed Solutions

1. Expand early voting
2. Liberalize absentee voting
3. Establish more convenient “vote centers”
4. Run all-mail elections
5. Institute an Election Day holiday
6. Schedule voting on the weekend
7. Hold elections online
8. Allow same day registration
9. Modernize (or eliminate) voter registration

Voting Technology
Funding to the states to purchase new voting machines provided by the 2002 Help America Vote Act (HAVA) prompted deployment of various voting technologies, ranging from optical scanners and direct record electronic (DRE) voting machines to centralized state voter registries to electronic poll books. Drafters of the statute and election administrators nationwide hoped that these technologies would prevent the problems seen during the 2000 presidential election. But technologies purchased and implemented in the wake of HAVA lacked continuing funding. With the downturn in the economy, as a general matter states have been unable to afford replacement machines leaving these states with technology that is now over a decade old and heavily used. Predictably, breakdowns are more common as machines age, leading to fewer functioning machines, and causing delays at polling places. Problems with voting technology culminated in voting delays in several states on November 6, 2012.

Addressing deficiencies in U.S. voting technologies has been widely discussed since HAVA's passage. Some studies have attempted to identify states with the most pressing voting technology challenges. The discussion below reviews voting technology problems from the November 6, 2012 election with an eye towards understanding the principal sources.

Recent data suggests that voting machines were less of a cause of delay in 2012 than previously assumed. Still, reporting on the 2012 election suggested three broad classes of technology used on Election Day impacted delays: DRE machines, optical scanners with paper ballots, and electronic poll books (EPBs) used at check in. Voters and election observers most frequently singled out DRE machines. In some states voter confidence may have played a part in the delays by leading voters to wait in line for their preferred technology instead of using a less preferred technology.

Each of these classes of technology faced unique and individual problems, though one problem appears common to all: an insufficient number of machines. Voter complaints about an inadequate number of voting machines were reported in 2004 and 2008, and some believe that insufficient machines decreased voter turnout in those elections. The inadequate number of machines may be due, at least partially, to budget constraints in recent years.

Summary of Proposed Solutions

1. Increase the number of available machines
2. Authorize the use of emergency paper ballots
3. Develop a National Clearinghouse of voting machine problems
4. Improve existing machines
5. Create an Enterprise fund dedicated to the purchase and maintenance of future voting machines
6. Negotiate Better Service Terms with Vendors or Penalize Non-Performance

Ballot Length
Ballot design first arose as a publicly visible election administration issue as the nation watched the infamous butterfly ballot disaster unfold in Palm Beach County, Florida in 2000. Much of the research since 2000 on ballot design has focused on the problem of consequent lost votes and voter confusion. Distinct from ballot design, ballot length has arisen as a prominent cause of election delay, particularly in recent years. This section will examine the extent to which ballot length caused delays on Election Day 2012.

In 2012, the length of ballots in several locations drew national attention as delays grew in the areas with the longest ballots. The idea that long ballots create long lines seems logical. The average person reads approximately 300 words per minute, meaning that a 4,000 word ballot will take the average person 13 minutes and 20 seconds to read. Parsing the complex language of propositions may take longer since comprehension limits maximum reading speed. Another suggestion is that recent delays may be due to an increase in the saliency of propositions. Some argue that increased use of electronic voting machines has led to a decline in roll off. Others posit the opposite effect, claiming that technology that uses a full-face ballot (such as DREs and lever voting machines) can increase the residual vote rate. The increased use of social media has allowed smaller groups to compete effectively against wealthier groups during initiative campaigns. This could lead to campaign activity that would not be feasible without social media, which could increase issue salience, and mitigate the impacts "choice fatigue."

Summary of Proposed Solutions

1. Increase voter education efforts
2. Increase the use of absentee or early voting
3. Shorten the ballot
4. Hold voting referendums and ballot propositions in non-presidential election years
    

Legal Instability
Legal instability immediately prior to elections causes voter and poll worker confusion, which can lead to delays. The 2012 election occurred immediately following the decennial redistricting, which in itself can cause confusion among voters about where to cast their vote. Explained one election official in Illinois, "There are instances of people going to the wrong polling places, as we expected\u2026. This is always the way it is after redistricting." To make matters worse, a slate of redistricting litigation delayed finality of newly drawn lines, leading to dramatically increased uncertainty in some states. Texas provides an extreme example. On June 24, 2011, the Texas Legislature passed its redistricting maps and submitted them to a federal court in the District of Columbia for preclearance under Voting Rights Act §5. The litigation forced a one-month delay of the primary election, which caused great confusion not only for voters, but for candidates as well. On August 28, 2012, the court refused to preclear each of the three maps. The delay (and eventual denial) forced the federal court in San Antonio to draw interim maps, which exacerbated confusion and led to delays on Election Day.

Strict photographic voter ID laws have gained much media attention over the last several years, and multiple states were party to lawsuits that produced murky results on what ID laws could be enforced and when. According to a representative for the Lawyers' Committee for Civil Rights Under Law, the majority of complaints her organization received on Election Day were related to confusion about ID requirements. Given the national attention focused on the voter ID debate, it is little wonder that poll workers and voters were confused about documentation necessary to vote. But states where voter ID litigation was brought witnessed the most reports of voter ID-related confusion, to wit Pennsylvania.

Students as a group faced increased issues with identification, too. Each state has different domicile requirements and some students are able to register using their dorm addresses with very little corroborating documentation. Each state also has different rules on whether college IDs can be used at the polls. Students may not have state identification readily available and feel uncertain about what they need and can present in order to vote, creating confusion. Whenever there is an identification problem, even if it is resolved by the voter digging through her wallet to find the correct document, it takes a little extra time to reach the voting booth.

Aside from confusion about redistricting and voter ID, changes to early voting in states like Ohio or disputes relating to changes to early voting in states like Florida in close proximity to the election are other examples of statutory instability that can contribute to delays. Each placed greater stress on the local election authorities when administering elections in their respective jurisdictions or confusion by voters. If 2012 is to be a lesson, legislators should take caution in changing voting laws too close to elections. Voters are easily confused by new rules, and poll workers need time to absorb new laws and train under new regimes.

Poll Workers
Although poll workers throughout the country rise to a difficult challenge and fill a needed role, human error and lapses in poll worker training affected wait times at the polls in 2012. The following list highlights some issues related to delays at the polls.

• Problems Processing Voters
• Limited Understanding of Technology
• Voter Interface
• Enforcing Polling Place Procedure
• Following the Schedule
• Not Enough Poll Workers

Summary of Proposed Solutions

1. Better or more training
2. More pay
3. Recruit younger poll workers
4. More accessible technical support
5. Modernized registration
    

Voter Confusion
A perennial problem in voting in the United States is poorly informed voters. At some level, some percentage of voters will always be misinformed; the goal however, should be to take steps to keep that number as low as possible. To this end, all states have voter education programs that vary considerably in nature. Some states post information on election websites and take little additional action to inform voters. Other states, particularly those with recent changes to voting laws, were more proactive in 2012. Below is a listing of categories in which voter confusion may have caused delays in 2012.

• Voters Do Not Know Their Polling Place
• Voters Do Not Understand Ballot Initiatives
• Voting Technology Confusion
• Language Minority Voters
• Misinformation
• Poll Watcher Disputes with Poll Workers Interference
• Challengers

Summary of Proposed Solutions

1. Improve voter outreach
2. Modernize registration
3. Improve language assistance
4. Improve poll worker training

Contingency Planning<\/u>

What happens if there is a hurricane on Election Day? When a national disaster threatens voter access, as Hurricane Sandy did for many on the East Coast in the 2012 election, public perception of electoral fairness is undermined. While the impact of Sandy on the presidential election was not considered to be outcome determinative, it is easy to imagine the turmoil had the storm hit just a week later. In the past twelve years, the United States has experienced at least three major disasters that threatened the integrity of voting on Election Day: elections in New York City disrupted by the September 11th attacks, displacement caused by Hurricane Katrina, and Superstorm Sandy. Despite these wake up calls, most states still lack a clear plan should a catastrophic emergency arise on or just before Election Day.