Yesterday I teased you about how big data, social media and psychographic profiles can be used to change the outcome of an election. Before we get into what the future holds, we first need to understand how things are done now and where we stand. As a reminder, I’m using the Norma Chavez, Veronica Escobar and Dori Fenenbock campaigns as an example. I’m also using the El Paso Votes APP for my big data source. Today, we’ll look at what the expected outcome is in the Congressional race and how campaigns get the vote out for their candidates.
I have seen or have heard about the results of a few internal polls among different campaigns and organizations, and what they predict the election will end like between Chavez, Escobar and Fenenbock. Polls are snapshots and not knowing the methodology and sample universe of the polls it is difficult to know how accurate they may be. In addition, internal polls are self-serving and if they are being shared with me, it is for an agenda. Donald Trump also proved that polls can be wrong.
However, they are a useful starting point.
There are three likely scenarios in the Chavez, Escobar and Fenenbock race. The first is an outright win by Veronica Escobar, squeaking by at just over 50% of the votes cast. Fenenbock finishes in second and Chavez finishes in a distant third-place. The second plausible scenario is a runoff between Escobar and Fenenbock because of Chavez taking votes away from Escobar. And, the third scenario is a Chavez-Escobar runoff.
According to the polling data I have seen, or heard about, and other conversations, Fenenbock has topped out about the 30-35% of the votes that will be cast. This leaves Escobar as the winner without a runoff. Where things differ in the information I have seen, is how many of the votes Chavez will take away from Escobar and whether it will be enough to force Escobar into a runoff.
One poll shows that Chavez takes votes away from Fenenbock and ends up in a runoff against Escobar. That scenario is unlikely.
The likely scenario is that Veronica Escobar squeaks by at just over 50% or is in a runoff against Fenenbock.
As such, for the purposes of our discussion we will look to see how big data, social media and psychographic profiles can be used by Norma Chavez to put her into the run off with either of the two candidates. Chavez is the underdog in the race and has little money and time to run a traditional campaign, so she needs to mount an untraditional campaign.
Remember that the techniques we will be discussing in the upcoming days should have been implemented earlier by Chavez to make any meaningful change in the outcome, but, nonetheless, as early voting starts next week, it is still possible, but unlikely that the techniques could put Chavez in the runoff.
Currently, and the likely get out the vote strategy being used by the three; Chavez, Escobar and Fenenbock campaigns is the traditional method of targeting likely voters. This is done by building lists of voters made up from the voter discs provided by the County.
The voter lists are likely the last three, or four primary and county-wide general elections. From these datasets, the campaigns create a master list of the “likely voter.” The Likely voter is the voter that cast a vote in the last three-to-four elections that the campaigns are using. Some campaigns create two, or three tiers: the super voter, which is the voter that cast a vote in each of the elections in the sample, and one or two additional tiers based on the number of votes cast by the voter.
Whether the source data is the county or the Democratic Party VAN system, the data is basically the same. From this data, the campaigns know who the voter is, how old they are, where they live and how many votes they cast over the period they selected. It is basic demographics information. Sometimes the data is augmented with additional data elements, like a telephone number or an email address. Some campaigns may even know if the voter is active in political groups, like the Tejano Democrats.
Those using the VAN system are limited by the inherent bias targeting Democratic Party votes that is part of VAN.
The resulting lists give the campaigns enough information to start a “ground game” targeting the voters on their lists from tier one, on down to what ever tiers they have decided to use. However, whether it is tier one, or one down, the target universe is large, thus limiting the messaging to messages targeted at wider audiences. For example, “I will work to reduce taxes.”
But what happens if, instead of a broad message to a larger audience, a targeted message specifically targeting a voter’s specific issue that consumes them at this moment was possible? For example, what if we knew that a frequent voter is a real estate agent. We would know that the voter’s livelihood is dependent on the number of homes they sell each year. Because this is a congressional race, local taxes are beyond the purview of the office, thus taxes, as one of the voter’s trigger points is not useful.
But, what about Fannie Mae and FHA mortgages?
Fannie Mae is a pseudo government agency dependent on Congress. Although it is considered a private enterprise, the mortgage company depends on the goodwill of Congress to continue to fund home loans. In El Paso, FHA, a federal government entity, is the predominant lender for home loans. FHA has less stringent requirements then traditional mortgages, especially when it comes to the down payment. Fannie Mae is required to provide loans when traditional lenders would not, because of Fannie Mae’s government linkages. Currently, Congress is looking at overhauling Fannie Mae because of the 2008 Credit Crisis. Congress also has say on how FHA handles private mortgage insurance (PMI) for first-time homeowners. PMI skyrocketed after the financial crisis.
So, instead of “I will reduce taxes for you,” the targeted message to the real estate agent was, “vote for me because I will work to bring down the burden of PMI”?
As a real estate agent, the likely voter would be amiable to a campaign messaging scheme telling him the candidate will ensure that Congress lowers the PMI and increases the lending capacity for both Fannie Mae and FHA. With these changes, the more homes the real estate agent has the potential to sell. In other words, the candidate is now talking to the voter’s trigger issue, instead of the macro issue of “lowering taxes.”
That’s the type of targeted messaging that this sample likely voter will react to.
That is just the tip of the iceberg. Tomorrow we will start delving deeper into building out, our get-out-the-vote lists for our example candidate. We’ll see that knowing the profession of the likely voter is just the beginning of targeted messaging for votes.