At the headquarters of Angela Merkel’s Christian Democratic Union (CDU) party, there is a big wooden door that goes nowhere and everywhere. Campaign manager Peter Tauber rings the doorbell and a man appears, barking “Who are you? What do you want?” Mr. Tauber calmly ends the conversation and the man disappears. The doorbell is rung again, and where once the angry man was, is now a young person who happily loads up with CDU swag. A third ring reveals a woman eager to talk to the party she supports and, with true German hospitality, invites Mr. Taber in for cake and coffee.
Except Mr. Tauber doesn’t actually get to enjoy the coffee break, but keeps ringing. The people he meets are not real people. In fact, they can’t even open the door for themselves because they are images on a screen behind the door – a simulation for door-to-door training exercises as Mr. Tauber readies his team for the 2017 federal election campaign. The door is in what Mr. Tauber calls the “boiler room” and gives campaigners a chance to practice their door-to-door strategies. But these won’t be blanket visits this year, but instead the targets will be defined with sought-out and paid-for data.
Data analysis in Germany is massively underdeveloped and not only with regards to political parties. Frank Pörschmann, Vice President of the Digital Analytics Association
In 2013, the CDU rode a sleeper car to victory under a strategy of “asymmetrical demobilization.” This year, Ms. Merkel’s incumbency is being squeezed from the Social Democrats (SPD) and the far-right Alternative for Germany (AfD). And all parties are using the same strategy: Potential analysis – using the publicly available data, including previous election results and regional socioeconomic data. Do people buy or rent their homes in the area? What do people earn? Do people subscribe to newspapers? The CDU is investing in both these numbers and their analyses to make their door-to-door strategy more effective, better targeting potential voters.
Across the aisle at the SPD, the strategy isn’t much different. “It’s about reaching out to people and providing the best possible support to our campaigners,” SPD campaign manager Katarina Barley told Handelsblatt. “Constituencies are being decided by increasingly tighter margins.”
All the parties are using their data analysis to support door-to-door efforts, both on- and offline. Mr. Tauber, together with Conrad Clemens, in charge of mobilizing the young party support under “Connect 17,” are looking to use what they call the “gorilla” of targeting, Facebook. Using the social network’s Facebook Elections platform, parties across Germany are spending anywhere from €1,000 to €8,000 for 400,000 to 3 million "placements," Facebook-speak for advertisements. It’s a strategy that received a lot of credit for President Donald Trump’s surprise victory and the unexpected results of the British referendum on the European Union membership. Companies like Cambridge Analytica are suddenly getting a lot of attention for the way American and British data were used to leverage voters to the polls.
There is certainly potential to use data more efficiently and effectively in Germany. In both the U.S. presidential elections and the E.U. referendum, campaigns used real-time data in order to determine how to best speak to potential voters on their own doorsteps. “Data analysis in Germany is massively underdeveloped and not only with regards to political parties,” states Frank Pörschmann, Vice President of the Digital Analytics Association.
In Germany, the amount of information available to parties, however, is limited. People here are weary of Big Bad Data and traditionally private, a lingering effect of the extensive personal espionage that happened in the 1930s and within East Germany. “Most of what we saw implemented in the U.S. election is, for good reason, not allowed here through privacy protection laws,” Ms. Barley noted. The law favors the individuals and political parties, for all their investment and spending, takes a conservative approach to their interpretation. For example, they’re not using data from Deutsche Post Direkt, a subsidiary of Deutsche Post that focuses on mail advertising and contact management with fully legal data sets pertaining to the socio-demographics, preferences and living environments of potential voters. “For us, it’d be wholly unacceptable to buy such data,” said Mark Seibert, campaign manager for the Left Party at the agency DiG/Plus. Robert Heinrich, campaign manager for the Green Party, reached a similar conclusion.
By its own account, the SPD uses only “publicly accessible data” in approaching its target group, which is to say it abstains from gathering personal data and voter preferences. Through the analysis of the data it does possess, the party seeks to gather broad information about a given electoral district and its voters.
The Green Party makes a “potential map” available to its candidates. “This is calculated according to secondary votes and voter participation in past elections,” said Michael Kellner, the Green Party’s federal manager. This is “the best indicator of future election results.” Members of the party can use the information to decide which doors they want to knock on. This strategy proved effective throughout the course of a recent local campaign in Berlin. “The focus is on the organization of door-to-door teams, the selection of the best streets, the interaction with people at their doors and also which election materials are most effective,” said Mr. Kellner.
Most of what we saw implemented in the U.S. election is, for good reason, not allowed here through privacy protection laws. Katarina Barley, SPD campaign manager
In the end, parties are looking at ways to best mobilize their campaign war chests while sticking to the country’s data privacy laws. The Free Democrats and Greens each have roughly €5 million at their disposal. The AfD has a campaign budget of €3.3 million. Both the CDU and SPD each plan to spend approximately €20 million. The parties are keeping mum, however, about how much of their budgets they plan to spend on data analysis.
There are already lessons learned in the first of three state elections, which took place in Saarland. There, the CDU’s campaign team recommended specific streets for their campaign workers. In the end, the party’s campaigners knocked on 75,000 doors and, for a quarter of those addresses, employed an app to make note of the locations, sympathies and political leanings of their inhabitants. This data was then sent to a database and transposed onto a map of Germany. The app allowed campaign workers to coordinate their activities and helped them identify CDU strongholds. This data, however, “only allows us to see the political mood on a given street,” Mr. Clemens explained. Still, the data has been retained for use in future elections.
Heike Anger is a Handeslblatt editor for economics and politics. Daniel Delhaes and Dana Heide are political correspondents working from Berlin's office with Silke Kersting, whose reporting focuses on consumer protection among other policies. To contact the authors: [email protected], [email protected], [email protected], [email protected].