Precinct Accessibility and Quality

Precinct Accessibility and Quality: A Field Experiment


The 2000 Presidential election highlighted the importance of polling place accessibility and quality with many voters not able to find their precinct or turned away once they arrived. A New York Times headline in November 2000 proclaimed “Arriving at Some Florida Voting Places, Some Blacks Found Frustration,” and the story concluded that problems such as inadequate parking, changing precinct locations, and uninformed poll workers. In the end, thousands of potential voters may have been deterred from voting on election day due to “low quality” precincts. What’s more, these lower quality precincts may not be randomly distributed throughout a city or state and instead might be concentrated in low-income and minority communities, which already witness lower rates of voting.

precinct graph1While many anecdotal stories surfaced in 2000 regarding polling place irregularities, no systematic evidence existed to fully assess the state of precinct quality. To investigate this issue, the Center for the Study of Los Angeles at Loyola Marymount University pioneered the first ever precinct accessibility study in March 2003. After designing a 30-item checklist of polling place characteristics, appearance and quality, student research teams were sent to randomly selected precincts throughout the city to record information about each one on election day. Since then, the project has been refined and will be conducted for the fourth time in Nov. 2004.

While the study has only been conducted in Los Angeles, the research has demonstrated that many inequalities exist with regard to precinct quality. To provide comparison sites, additional cities across the country should be examined to determine whether or not voters face obstacles at the polling place, or they can vote with ease. Other major American cities such as New York, Chicago, Houston, and Miami should be incorporated for a more accurate picture of voting equality. While these major metropolitan areas offer a mix of ethnic and class diversity, smaller rural populations could also be studied to further compare precinct characteristics in more homogenous populations. The study is fairly simple to setup and requires little more than student researchers to carry out.


How to Get Started
The first step is to contact the County Registrar of Voters or local election division and obtain the complete list of precinct sites and addresses for the upcoming election. In addition, request the precinct-to-census tract match list which will allow you to overlay each precinct with census bureau demographics at the census tract level. This should be done about 3 months in advance, although county officials may still be finalizing the list as early as two weeks prior to the election. The next step is to determine how many student teams will participate. Students should be paired in teams of two, where one of the two students has a reliable car. If you have 50 student participants, you can create 25 teams, and you divide the city/county into 25 equal portions. This will ensure that the entire city is covered and the sample is not biased. Typically, you can sort the precinct address list by zip code, which are geographically approximate. If you have 1000 precinct addresses, and 25 teams, you will divide the full list into 25 equal portions (sets) each containing 40 precinct sites and addresses. Within each “set” randomize the precinct order by sorting on the zip code variable ascending in half and descending in half. In our prior experience, student teams have been able to complete on average 20 precincts, however some teams may complete 30. It is important to establish a minimum requirement of 20 precincts visited so that the sample is not geographically biased.

precinct graph2

After the student teams have been assigned, conduct one day of training (typically during class session about three weeks before the election) to introduce the students to the project, explain the checklist, and stress the importance of valid data. At the end of the first training session, ask students to familiarize themselves with zip codes in the city and make requests for 5 zip codes in the city they would like to visit. While you cannot guarantee that students will receive precincts in those zip codes, we have found that students are more efficient in parts of town they are more familiar with. In addition, it more closely resembles the voting experience of the median voter on election day, who is likely to be somewhat or very familiar with the neighborhood surrounding the polling place. Once precinct assignments are made (by zip code) distribute the list of 40 precinct sites and addresses to the student teams (about two weeks before the election) and ask them to map out directions to each precinct using Mapquest, Yahoo! Maps, or Thomas Guide. Each student team should be identifiable by the last names of the students and also a team number.

On the final class session before the election, conduct an additional training session to answer final questions, explain the meeting time and location for Election Day, and again stress the importance of reliable data. To ensure that students visit every precinct that they complete a checklist for, they are required to take a photograph, using a Kodak or Fuji disposable camera (with 27 exposures) of the precinct. While most of the questions on the checklist are straightforward and objective, it is important that the internal coding is consistent. To address this potential shortcoming, make sure that there are two precincts that overlap from one team to the next. At the end of the project, select only these precincts which will have been visited twice, and compare the checklist responses of the different student teams. Using various test and reliability statistics, you can assess the coding consistency of the precincts in the overall study.

Finally, on election day, ask the students to assemble at 8:00 AM in an auditorium or large room on campus. If possible, request that the students wear a shirt or hat identifying the university. The main purpose of the final meeting on election day is to distribute the packet of information and tools, and answer any remaining questions. Each student team should receive:
  • Clipboard and Pencil
  • 50 Front/Back Copies of the Checklist
  • Basic City Map
  • Disposable Camera
  • Box Lunch
  • Unique Precinct Address List for Team
  • Voter Bill of Rights
  • Letter of Notice from Department or University President

The Voter Bill of Rights should be available from the Secretary of State and is likely posted on the state website. This is a basic overview of voting rights and responsibilities, and most importantly, it includes a provision stating that independent monitors have a right to visit polling places on election day, make observations, and ask questions of poll workers. It is important that students are not turned away from precincts, but if they are, ask them to record as much information as possible. The Letter of Notice should be a very brief letter printed on University letterhead and written by yourself, the department chair, or the University President that is addressed to the head poll worker, and outlines the scope of the project in the case that the head poll worker won’t allow the students into the precinct.

Finally, after the students have conducted the study and returned the completed checklists, enter the data into the Microsoft Excel using the accompanying codebook and template. Export the data into SPSS, Stata, or a similar statistics package to run basic frequencies and report the results back to the students. While most of the precincts that one team visited might have been above average, they will be surprised to learn that throughout the city many differences existed and not all precincts are equally accessible.