Speak to a Human: 410-205-5478

0 Item Daily: 6AM - 9PM EST

Don't Throw the Baby Out With the Bath Water, re-Define Super Voters.

Posted by Shawn Harmon on 4/24/2015 to Voting

Nearly every campaign that contacts us requests a list of "Super Voters" also known as "Quantum Voters" aka "Perfect Voters" aka "Likely Voters".

The premise for selecting Super Voters is threefold. What campaign is not constrained by finite dollars, warm bodies and the clock? And to tailor messages that are more relevant to an active voter audience as opposed to those who vote less frequently is arguably an effective maximization of budget too, right?  

Campaigns have requested that eMerges provide them variations of Super Voters including but not limited to voters who "voted in the last three generals" or a bit more obtusely,  voters who  "voted in the last three local primaries but not the general election of 2012". 

Political professionals recognize that voting history is more nuanced than just grabbing all the voters based on their past voting history---sometimes the environment surrounding a particular election matters too. Super Voters are often selected based on these additional variables:

a) Truly contested elections
b) Non-federal versus federal year elections
c) Presidential versus non presidential years
d) Month of election
e) County versus state senate races
f) Closed versus open primaries
g) Primary versus general elections

You get the idea. The specific relevance  reflected for each historical election  depends on the dynamics of your race. A smartly run campaign is assumed to consider the quirks of each election when customizing its list of Super Voters.  

Although voting history is not the only filter used to create Super Voters, voting history always seems to be at least one of the criteria used for targeting. Depending on what the candidates are arguing about, Super Voters are often coupled with other voter filters including a blend of modeled, self-reported and third-party data such as: race,  party, gender,  carry concealed weapon permit holders, registration date, weed legalization supporters,  age, low taxes, marriage equality, income, education attainment, and of course Online activities.  eMerges offers about 100 other demographics but the majority of our customers usually seek 3-4 filters with which to customize their voter lists.

More often than not, campaigns contacting eMerges  approach us already bound and determined to acquire a list of Super Voters only. The wiliest campaign managers however do not just ask us for a list of Super Voters. Instead, they also seek newly registered voters. This is because a percentage of newly registered voters go on to prove themselves to be Super Voters and these managers are on to it.

How many newly registered voters become Super Voters and what is the consequence of ignoring newly registered voters by focusing only on Super Voters? We took a look at three counties to get some insight.

1) Anne Arundel County in Maryland, 
2) Fulton County in Georgia and 
3) Orange County in California

First we wanted to identify newly registered voters. We selected anyone who registered between October 1, 2002 and November 1, 2006. Why this range? We were selecting Super Voters for an election in November 2007, the people who had registered between 2002-2006 would have been excluded from our 2007 Super Voter list because they were newly registered voters. These voters were then watched to see how many actually become Super Voters by voting in each of the elections 2008, 2010 and 2012.

How likely is a group of these newly registered voters---whom we categorically ignored in 2007---to subsequently become Super Voters by 2013?

Anne Arundel County, Maryland 
Registered Voters Lists

Orange County, California
Registered Voters Lists
Fulton County, Georgia
Look at how similar Figure A and Figure B are for each of the three counties, you can plainly see that in each county:

New voters become Super Voters at the same rate as all voters.

 Between Figure a and Figure b, there are equal proportions of Super Voters for all county voters versus the new voters. For example, in Anne Arundel County, Maryland 48% of all voters were Super Voters and 48% (Figure a) of newly registered voters were Super Voters too (Figure b).

All three sets of county pie charts prove that newly registered voters from 2002-2006 become Super Voters from 2008 - 2012 at about the same rate as all the other county voters.

So what were the consequences to a campaign from back in 2007 which selected Super Voters only? Using additional 2002, 2004 and 2006 voter data the results for ignoring new voters was perilous. Huge chunks of new voters went on to become Super Voters. Campaigns which targeted only Super Voters ignored 26+% of the voters who actually voted.

A campaign may believe that they are saving scarce dollars by targeting Super Voters only but the campaign may not realize how many future Super Voters are hidden in the newly registered voters list. To reach the newly registered Super Voters of 2007 a campaign would have to have targeted all newly registered voters from 2002-2006. This would have meant incurring the cost of the subsequent  processing  of voter mailings, emails, phone calling or FaceBook fees---in Anne Arundel County, 58,188 voter records would have been  required just to reach the 28,171 who would later prove to become Super Voters.  

Any campaign could select Super Voters using less stringent criteria, for instance by selecting anyone who voted in 2 of the 3 last general elections. The problem with ignoring the new registrants is just minimized though. Targeting newly registered voters carries its own challenges--are they young or just new to the area? You will likely just not know as much about them. On the other hand, what if yours is the only campaign that reaches out to future Super Voters? 

Pretty obviously, targeting Super Voters and newly registered voters while also harnessing the knowledge of political party and other quirky audience profiles will enable any campaign to avoid throwing mud up against the wall in the desperate hope that some actually sticks.

There is a lot of marketing hype and hoopla with Super Voters. Our customers report  that other registered voter list sales people pitch Super Voter targeting as the irrefutable Holy Grail and some campaigns buy into that sales pitch hook, line and sinker. 

>Summary: new voters become Super Voters at the same rate as all voters. Campaigns which targeted only Super Voters ignored 26+% of the voters who actually voted.<

Don't throw the baby out with the bath water---Super Voters are not enough. Make an evidence based decision. The statistics indicate you should target both Super Voters as well as the newly registered voters. The relatively small cost of processing the extra voter records will surely reduce your risk of loss, right?

*Regarding the Data - Clearly the nuances that affect elections need to be included in a deeper analysis. Factors affecting the outcome: date ranges we selected, election dates, registration drives and close dates, HAVA updates, in county re-registrations, blends of modeled or self-reported and third-party data---are all variables that should be included in further analysis. If you would like to see the original data employed, email Data@eMerges.com.