Making Data Work for Your Campaign
By Jeff McGowan
Data is the most rapidly expanding driver of market growth and industrial evolution. The size of digital data is expected to double every year between 2010 and 2020, growing 50-fold in that time period alone. As the amount of available data grows exponentially, our ability to analyze that data has lagged behind. Only an estimated 0.5% of available data ever gets analyzed.
While big data analytics was quickly adapted by Fortune 500 companies and the private sector in general, political campaigns always seemed to lag behind. With recent investments by national party committees in big data, political campaigns now have access to more data than ever before. But access to robust data sets doesn’t do a lot of good just sitting in data repositories.
While campaigns have access to incredibly useful data tools like modeling and voter scoring, these tools can come at costs that price out all but statewide and the highest tier Congressional races.
But data can be made available at little to no cost, either directly from the local or State Board of Elections, or from the State or National Party Committee. From there, data analytics can be run without a PhD level data scientist on hand, although it certainly doesn’t hurt.
Below are some examples of data sets that are generally publicly available:
• Current Voter File – In most jurisdictions, lists of registered voters are a matter of public record. This usually includes the majority of the fields provided on their voter registration form, such as:
· Phone Number
· Party Affiliation (where applicable)
· Political Geography (City Council, State Assembly/State Senate, Congressional, etc).
· Previous Vote History
• Early/Absentee Data – Where applicable, lists of requested absentee ballots and their return status, as well as lists of voters that have already cast ballots can be made available to campaigns. This is a great way to identify voters interested in the election that have taken the step of requesting a mail-ballot. This list should also be used to scrub out people that have already voted, thus reducing your voter contact budget and eliminating waste.
• Precinct Level Election Results – Previous election results at the precinct level. Useful data set in building vote goals based on either your own previous performance or the performance of other candidates.
While getting your hands on the above data may be easier (and cheaper) than you think, you still have to be able to work with these data sets to both gain insights and make decisions.
Again, statistical modeling and voter scoring is a great resource for those that can afford it. But, plenty of work can be done with just a cursory glance of the voters in your file or of the previous election results in a given jurisdiction.
• Run basic counts on your voter file
· How many Republicans, Democrats, and Independents are in your district?
· How many of them voted in the last election?
· Based on last election’s turnout, how many voters from each party would you need to win?
• Evaluate precinct level data to glean insights
· What is the average Republican candidate performance in each precinct?
· Which precincts can I get the most amount of votes out of?
Again, high powered data solutions are always great, for those that can afford them but basic data investigation can be done at almost no cost at the beginning of a campaign.
While plenty of large political data firms exist to do data analytics for million dollar campaigns, product offerings from Bedrock Data Solutions can be scaled down to provide insight, targeting, and analysis to lower budget campaigns.