Will Data Win The US Election?

If so, Donald Trump is shooting himself in the foot


Hillary Clinton versus Donald Trump. It is a choice that many in the US are unlikely to be pleased with. One started as a joke and became a dangerous reality, while the other went from being viewed as a President-in-waiting to untrusted and corrupt. Few would say that it has been boring though, and as we move towards the November election we will likely see considerably more fireworks from both.

Although many will be looking to the debates as the key battle ground, it is unlikely to be what wins out. This is going to be the use of data.

At present this is a battle that is being resoundingly won by Clinton, who has appointed Teddy Goff as chief digital strategist, Elan Kriegel as analytics director, and Andrew Bleeker as an outside adviser. Donald Trump on the other hand, only recently hired a Pollster to his staff and declared analytics to be 'overrated'. This lack of concentration on data could come back to haunt him in a big way though, with recent elections showing that the use of data has been a significant factor in the eventual winner's campaign.

Data is incredibly valuable as it gives team the opportunity of micro-targeting - allowing different messages to go out to different areas of the population depending on how they are likely to react to it. For instance, a message sent to a hispanic man over 40 with 3 children will be different from a white woman under 30 with no children.

At this early stage of the campaign, such data is collected broadly, often through simple surveys, with questions like ‘Who are you likely to vote for' and 'What do you think of this candidate'. This allows pollsters and data scientists to collect information that enables broad groupings which can then be elaborated on with other data, such as voting records, campaign volunteering history and the like. This allows the data teams to create additional segmentation around their target voters, increasing the focus that can be put onto individual groups.

As we as a society create more and more data, an increasing amount of this is available to a candidate's data teams to paint an even deeper picture of their voters' wants and needs. For instance, if you 'Like' a candidate on Facebook the candidate's data team get access to a considerable amount of information about you, which can then be used to target messages, find pockets of voters and craft effective messaging. This is one of the elements that is so confusing about Donald Trump's campaign. He has over 7 million likes, compared to Clinton's 3.6 million. Not taking advantage of this is like fighting with one hand behind his back.

His lack of data use is not only potentially damaging to his chances of getting to the White House in 2016, it may also significantly damage the Republican party moving forward. This is because data in elections is not a new thing, it has had significant importance in elections for decades.

The 1964 election was particularly important in that it set the Republican platform that has seen them increase their standing as the 'conservative' US political party. In this campaign the data wasn't the same as today, but more a case of simple geographical profiling, collected by the team for the Republican candidate Barry Goldwater, who eventually lost to Democrat Lyndon Johnson. He may have lost (and most bookmakers still have Donald Trump at significantly longer odds than Hillary Clinton) but his emphasis on data collection saw Republicans elected in 5 of the next 8 elected Presidential elections. It allowed them to target specific conservative leaning areas that they previously didn't know about and hence build strong campaigns in the future.

It is something that both Obama campaigns have done during the last two election cycles on the way to becoming the first president in 50 years to win over 51% of the popular vote in two consecutive elections. The work done by his data team has created the most comprehensive voter data set ever assembled, which, once the nomination is confirmed, will give Clinton a huge advantage over her rival. If this data set can be even further improved during this election and the Trump campaign chooses to largely ignore theirs, it will have a significant impact in future elections.

With the data we have access to today, candidates have the opportunity to go considerably deeper than Goldwater's team and potentially Obama's too. With the success that has been seen using data, Donald Trump's stance could be a real boon to the considerably more data-savvy Democrat team. 

Bean small

Read next:

City of Chicago: An Analytics-Driven City