Big Data and Voter Registries – Election Officials

From analyzing the migratory track of the Canada Goose and its effects on the number of bird/plane interactions to the predictive modeling and profiling of consumer purchasing behavior.The use of big data has become prevalent in today’s society. Tracking voters intentions is no different. 

There has been a large increase in the number of data mining/farming firms establishing themselves in the use of big data in the electioneering fields: 270 Strategies Inc., NGP VAN, Goddard Gunster Inc., Elections Impact Group and Campaign Communications Group just to name a few. The application of big data in the election process extends over both the area of campaigning and election administration of voter list management and poll GIS.

Today we are looking at the application of big data in the maintenance and creation of voter registries. We know from statistics, that there is on average over 20% movement in the general population each year. There are numerous methods that jurisdictions use to maintain their voters list. Some jurisdictions require voters to register themselves. Others maintain their own list with the use of data from sources such as vital statistics (birth, death and marriage certificates), DMV records, tax records, postal NCOA (National change of address program), returned mail, assessment records, personal property records, property tax records, correctional institutions, health institutions, new citizenship records, Foreign Affairs(citizens residing outside the country), armed forces,  and the list goes on.  There has even been a sharing of information between multiple levels of jurisdictions, from federal to province/state to municipal levels.

There are numerous companies that have developed software for the management of these lists over different platforms including: SEO, SCYTL, Datafix, KVM*, Dominion Voting and many more. Each with its own strengths and weaknesses.

Summary of countries that maintain a voter registry and some of the methods used to compile the registers **

Links to national population records	Albania, Belgium, Botswana, Canada, Croatia, Denmark, Finland, Hungary, Iceland, Italy, Lithuania, Mexico, Moldova, Netherlands, Norway, Pakistan, Panama, Poland, Seychelles, Slovakia, Spain, Sweden, Ukraine
Links to police records of residence	Armenia, Austria, Belarus, Croatia, Czech Republic, Lithuania, the former Yugoslav Republic of Macedonia, Slovakia, Switzerland
Links to applications for government services	Australia, Canada, Cape Verde, Chile, Japan, Panama, Seychelles, Slovakia
Registration by voters at registration offices	Australia, Bahamas, Barbados, Bosnia and Herzegovina, Botswana, Burkina Faso, Cambodia, Cape Verde, Chile, Costa Rica, Croatia, Dominica, Dominican Republic, El Salvador, France, Germany, Guatemala, Honduras, Ireland, Lesotho, the former Yugoslav Republic of Macedonia, Mexico, Moldova, Mozambique, Namibia, Nicaragua, Pakistan, Panama, Paraguay, Portugal, Saint Lucia, Saint Vincent and the Grenadines
Door-to-door registration campaign	Albania, Australia, Barbados, Belarus, Costa Rica, India, Ireland, Mexico, Pakistan, Panama, Saint Vincent and the Grenadines, Seychelles, South Africa, Turkey, United Kingdom
Registration by mail	Australia, Bosnia and Herzegovina, Canada, Germany, Ireland, United Kingdom;
Mobile election registrar	Australia, Mozambique, Namibia, Panama, Uganda
Internet registration	Australia, Canada, Denmark

One of the largest issues facing data professionals in the maintenance and use of outside lists to update registries is the non-standardization of data sets across numerous files. Each contributing organization has its own record layout standards based on each organization requirements.  To complicate this job there is a continuous updating of municipal GIS systems to keep up with street name changes for 911 services, new residential development and redefining of polling district due to population movement/increases.

There are a number of tools, both open source and third party licensed options for dealing with standardizing the maintenance of your RDBMS for any platform, whether the platform is Oracle, SQL, DB2, NoSQL or any other. One free service is “Google Refine” that provides an easy way to clean up messy data. Another open source solution you may want to review, that is part of the “big data” craze, is “Hadoop “or “SQOOP” and their third party suppliers, Hive, Flume, Haze etc. If you are not comfortable with or find the open source solutions difficult to use then there are commercial versions of Hadoop including Cloudera with a more user-friendly feel.

With the advances and ever-changing tools available for big data, IT managers have access to data manipulation and analytic tools on their desktop that was only available to larger organizations. What new tools will be available next?

**Source: The Election Process Information Collection, a joint project of the Institute for Democracy and Electoral Assistance, United Nations Development Programme, and International Foundation for Election Systems (www.epicproject.org).