The quality of Election Administration and Voting Survey (EAVS) has gotten better over time, and more organizations and individuals are pressing for data-driven election policy than ever before, which bodes well for the role election data plays. We hope to see in the near future wider access to election data and greater public understanding of what it tells us about how we pick our leaders. One great way to accomplish both these objectives is through data visualization, which a number of election offices across the country have started to practice. We've included an example here, a visualization of the 2014 EAVS data on overseas civilian and military voters.
August 2015 saw the first-ever Election Assistance Commission (EAC) election data conference, a long overdue gathering to discuss all aspects of the election data collection and dissemination we see in the U.S. The EAC has been collecting data from counties and municipalities in all 50 states and territories for a decade. Perhaps the most-discussed topic at the conference was that of data quality: what works well and what is needed to improve EAC's data collection, which takes the form of the post-election EAVS.
It's no secret we administer elections in myriad ways, and data collection is no different. We have no common electoral vernacular, so local election officials interpret the questions in the EAVS and other survey instruments in different ways. Budgetary, logistical, and technological capacities vary from election office to election office, and the time and effort officials can dedicate to working with data vary with them. State and local laws also impact how data is collected and reported. As a result of these factors, data completeness and standardization have emerged as pressing issues in data quality. It's difficult to understate the importance high-quality data has for our electoral system. Data is necessary to effectively allocate resources. Data is necessary to gauge the quality of the voting experience. Data is necessary to measure the impact laws and regulations have on the electoral process. Data is necessary to keep the discussion of election administration grounded in reality and not politics. And most importantly, data is necessary for the evidence-based policymaking necessary to make our elections run better.