Home> How to win an election campaign

27.04.2017

How to win an election campaign

Every week, Arthur Muller and Guillaume Liégey teach Sciences Po students the best methods to help a candidate win an election. Those expecting to meet a pair of spin doctors who'll teach them tactics worthy of the series "House of Cards" will be disappointed. Our two lecturers, co-founders of a start-up specialised in electoral data, prefer to use the latest research and big data to develop their strategies. Interview.

Your course is called "Campaign Strategies: How to Win an Election"(Fr). Does this mean you have a miracle solution for getting a candidate elected?

Arthur Muller: Of course not! The course is essentially practical, even if we flesh it out with the latest research on electoral tactics. The course is presented as a role-play that lasts the whole semester. Each student is given a party and a constituency, and must try and get his or her candidate elected. Our role is to help the students identify the key phases of an election campaign. There's the electoral strategy, for example: looking at past election results and the socio-demographic profiles of the constituencies, and seeing who votes can be taken from. Public speaking exercises are another important aspect of the course. The students have to make a campaign speech in front of the others, or simulate a press conference where one student takes the role of the candidate and the others play the journalists asking the questions. Finally, there is also the activist side of things that's very important in winning a campaign. One of the exercises we propose is to make a video to rally activists for the final leg of a campaign.

Who are your students and what are they looking for in your course? Are they already involved in political parties?

A.M.: About half of them are already involved in campaigns but—this is going to surprise you—it’s not in the spirit of the course to talk politics! We assign our students a constituency and a candidate at the beginning of the semester by drawing lots; the students don't necessarily get the party they sympathise with. The bottom line is that we're here to address questions of strategy with as rational and scientific an approach as possible. That's the difference between an activist, who will in essence be partisan, and a strategist, who is there to think out a campaign. People often think it's all done intuitively. There may be some degree of intuition inherent in this profession, but in our course we try to bring more concrete factors to build on: data, experimental psychology and so on.

You’re the founder of a start up, Liégey Muller Pons, which uses big data, open data and research to strengthen campaigns. In your opinion, are these the tools of the future?

A.M. : Undoubtedly. Our inspiration is the American presidential campaign of the 2000s, which saw a major technological leap in how to map and cover an electoral area. The app developed by our company, "Fifty plus one", collects electoral data and collates it with INSEE data to obtain about a hundred variables and a very detailed analysis of the electorate by constituency. In 2012, during the previous French presidential campaign, this type of tool was often still seen as a gadget, but today almost all candidates have apps for mapping and activist purposes. Moreover, when our former students join campaign teams, it’s usually precisely to manage these aspects of a campaign. So they must have the right skills!

Are the traditional strategies for reaching voters like canvassing, posters, etc. outdated? 

A.M.: Our belief is that good election campaign communication is a subtle blend of “human” and data. It’s generally that balance that is most effective. American research has shown, for instance, that door-knocking remains one of the best techniques for reaching an electorate, although it was thought to be outdated. Human contact remains essential, but to be as effective as possible, it must also be based on data-driven tools to target the electorate in the right way.
 
Find out more