Stewart.Managing.Polling.Places

[Image via electionupdates]

My friend and colleague Charles Stewart – an MIT political scientist who is also a high-octane electiongeek – has a new report produced as part of the CalTech/MIT Voting Technology Project entitled “Managing Polling Place Resources“. Here’s what he has to say in a blog post at ElectionUpdates:

Just as the one-year count-down for the 2016 presidential election has begun, the Caltech/MIT Voting Technology Project has released a new report today about managing polling place resources.  Click here for the executive summary, and here for the full report. [emphasis added]

This report serves as a companion to a set of Web-based tools that the VTP developed and posted at the request of the bipartisan Presidential Commission on Election Administration (PCEA), to facilitate the recommendation that local jurisdictions “develop models and tools to assist them in effectively allocating resources across polling places.”

The report takes several new steps in the effort to spread the word about the usefulness of applying queuing theory to improve polling place practices.  First, it provides a single source of facts about lines at polling places in 2012 (with some updating to 2014).  Second, it provides a brief, intuitive introduction to queuing theory as applied to polling places — with a brief list of suggested readings for those who would like to learn more.  Finally, the report uses data from two actual local election jurisdictions and walks through “what-if analyses” that rely on the application of the resource allocation tools.

The report released today provides basic facts about where long lines were experienced in 2012 and which voters — based on race, voting mode, and residence — waited longer than others.  Information about the 2014 election updates previous research, and underscores how long lines tend to be more prevalent in on-year (presidential) elections than in midterm elections.  Beyond providing basic facts about the location of lines in American elections, the report provides a basic introduction to the science of line management, queuing theory, and a list of further readings for those who wish to delve more deeply into the subject.  Finally, this report demonstrates how the Web-based tools might be used, by working through actual data from two local jurisdictions.

The report is part of the Polling Place of the Future Project (PPOTF) of the VTP, which has been generously supported by the Democracy Fund.  Since the release of the PCEA report, the VTP calculator website has been visited thousands of times by users across the country (and around the world.)  We have received feedback from numerous jurisdictions about the utility of these calculators, as state and local officials try to effectively allocate their limited resources.

In recent months, two of the resource calculators have been updated, and those updates have been posted on the site.  The new versions include improvements to the user interfaces and the ability to upload data from multiple precincts, which allows the simultaneous analysis of hundreds of polling places for large jurisdictions.

With the one-year countdown to Election Day 2016 already underway, some might say that it is too late to make use of such analytical tools to make a difference in the next presidential election.  However, my experience is that most election administrators are always looking for ways to improve the experience for voters; thus the publication of a report that highlights how existing tools might help them prepare for November 2016 comes at the right time for those election administrators who are looking to fine-tune their plans for next year.

The section with the real-world locations (names changed to Metro City and Magnolia County) is especially fascinating, as it demonstrates how, based on observed line data, the addition or subtraction of resources like voter check-in stations can have a dramatic positive (or negative) effect on wait times and the voting experience. It’s also very honest about how adding resources is neither without cost nor independent of other factors like voter ID rules. Consider this excerpt that concludes the discussion about suburban Magnolia County:

Reducing check-in times would not be a trivial task. Much more goes on when a voter checks-in in Magnolia County than simply checking their name off an electronic list. Magnolia County is in a voter ID state, so the ID needs to be verified before voting, which is not true for Metro City. (However, Magnolia County is able to read information off of a voter’s driver’s license electronically, which should speed up the process.) When they check-in, voters are asked if their addresses are up-to-date in Magnolia County, which is not the case in Metro City. These additional time-consuming tasks may be mandated by the state, or may pay off in other ways.

Magnolia County has no clear path to reducing wait times to check in. Adding more check-in stations would impose a serious financial challenge to the county [because the county uses electronic pollbooks]. Cutting the amount of time to check-in would involve more than simply talking faster, but would require a thorough review of administrative practices and a revamping of training. However, despite the fact that substantially reducing check-in times in Magnolia County would be expensive, the use of resource allocation tools gives the county something to aim for, and makes the case for any additional resource needs. [p. 30]

This report is the next step in the effort to educate election officials and policymakers/budget-writers about the power of using observed data and analytical tools to prepare for Election Day. Charles and the team at the Polling Place of the Future Project are pushing the field forward in a crucial arena of importance to communities nationwide. Thanks to the Democracy Fund for their continued support of this work – but especially to Charles Stewart, who is himself a valuable (and for now, seemingly limitless) resource for the field.

I look forward to seeing how (and how well) jurisdictions apply these lessons in 2016 and beyond … stay tuned!