On a June morning, an giant blue and red “C” logo was displayed in front of the Federal Press Office in Berlin, located on the Spree River in the heart of the capital. It was essentially the German government screaming for attention, and why not? Germany’s flagship project in the fight against the coronavirus pandemic was ready for prime time. Finally. “#Ichappmit,” a billboard read. “I’m using the app.”
Inside, five representatives of the German government and two board members from Deutsche Telekom and the software company SAP were on the stage, along with the president of Germany’s center for disease control, the Robert Koch Institute. They could easily have been mistaken for happy parents after a difficult birth.
Helga Braun, the head of Angela Merkel’s Chancellery and a member of her conservative Christian Democratic Union (CDU), said that although it wasn’t the first, it was perhaps the “best” corona app available worldwide. Interior Minister Horst Seehofer, likewise of the conservatives, praised the “first class” experts in the ministries. The CEO of Deutsche Telekom enthused that the app was a “rock star.”
The only person who seemed to be trying to manage expectations was German Health Minister Jens Spahn of the CDU. The app, he said, is “no panacea.” The sentence was a bit like the rip cord on a parachute. If expectations are kept low, you won’t hit the ground as hard if things go wrong.
That moment of euphoria was almost 100 days ago. Since then, more than 18 million people have downloaded the app, and it has been rated on the Apple and Google app stores with 4.4 and 3.1 out of five stars. The German government, for its part, considers it to be a great success.
“One Tool Among Many”
In fact, though, hopes that the virus might be contained using the app have given way to disillusionment. There is no longer any talk of the “very central building block” of pandemic management, as government spokesman Steffen Seibert described it even before the app’s launch. It has since become “one tool among many.”
But what has the app actually achieved? To what extent is it helping contain the pandemic?
It’s difficult to issue any kind of interim progress report due to a lack of solid data available. No one can say, for example, exactly how many smartphone users are actively using the app. “We currently assume that figure to be 14 million,” Deutsche Telekom responded when asked. The German Health Ministry, meanwhile, answered the same question with 17 to 18 million.
Comments on the internet suggest that many users are annoyed by the app and its strange error messages or confusing warnings. One customer recently asked in the App Store what the new error message “EN_Error” meant. One developer replied that it was an Apple problem. “Neither a reinstall nor restarting” the app would help, the developer wrote, but “sometimes the errors go away by themselves.”
The number of glitches has been painful, particularly because the app was hardly a bargain. It cost the government 15 million euros to develop it, and a further 44.4 million euros have been earmarked for “maintenance and care” during the operation of the app through 2021. As of August 27, the government had also spent 9.4 million euros advertising the app. “This app deserves your trust,” Chancellor Angela Merkel ensured her podcast listeners shortly after the launch.
Despite all the advertising, the numbers of downloads have been growing only slowly for some time now. Every second user now considers the app to be ineffective, as a survey for the initiative D21 and the Technical University of Munich has shown.
Further Development Needed
Some health and digital experts are urging for the app to be revised as soon as possible before infection rates start to increase even more with the arrival of the cold season in Germany. “The app urgently needs to be further developed in order to make it effective,” says Karl Lauterbach, the point man for health care policy for the center-left Social Democrats (SPD) and an epidemiologist by training.
Manuel Höferlin, the point man for digital policy in the parliamentary group of the business-friendly Free Democrats (FDP), is critical of the government for having “rested for too long” following the app’s successful launch. With the exception of troubleshooting, he says, nothing has happened since then. He says it is “completely incomprehensible” to him why the app hasn’t been made available since then for older phone models and in App stores for people under the age of 17, as well.
The FDP politician also accuses Health Minister Spahn of having sowed confusion among users himself. “At times, he talked about the corona warning app and the data donation app and a quarantine app at the same time, which made many people uncertain and sacrificed important trust,” he says.
The truth, though, is that it was Spahn himself who got the app project rolling. It seemed to fit in perfectly with his agenda. When he became health minister two years ago, Spahn told an all-staff meeting at the Health Ministry that digitalization would be one of his core focuses. To back up that commitment, he set up a separate department for digitization, headed by Berlin-based health policy and digital health expert Gottfried Ludewig.
In the corona crisis, the Warn-App is one of the few measures with which he can stand out, given the federalist ramifications of a health policy system in which much of the responsibility is held by states and local governments. Early on, Spahn pointed to countries like South Korea, which succeeded in using mobile phone data to stop chains of infection. Spahn also saw the app as a way to get out of the lockdown in the long run.
But Spahn was too brash when it came to the implementation of the project, which unsettled many people. Initially, he wanted to enable the health authorities to request mobile phone cell data from telecommunications providers to trace infection chains. Following fierce protests, including objections from the Justice Ministry, Spahn backed down.
In the end, he pleaded for risk assessment to be carried out centrally on a server maintained by the Robert Koch Institute to obtain more data for pandemic control. That, in turn, also triggered protests. Hundreds of scientists and experts warned of “unprecedented surveillance.”
A Watered-Down App
Ultimately, the German government opted for what is called the decentralized solution – one in which the risk of coronavirus infection is determined by the smartphone itself. Apple and Google had announced that their operating systems would only support decentralized variants anyway.
Almost overnight, Chancellery Chief of Staff Helge Braun and Spahn commissioned the heads of Deutsche Telekom and SAP to develop that variant. It now meets the requirements of privacy and data protection, but it also weakens the app’s central task: that of stopping chains of infection early and widely.
At least that’s the view taken by Patrick Larscheid. The physician is the head of the public health department in the Berlin district of Reinickendorf. Each day, his team what the app should be able to take care of on its own: They perform contact tracing to warn people who may have been exposed to the coronavirus. They go about their work using traditional means – by interviewing infected persons and then phoning their contacts.
Theoretically, the app should be able to make such work easier. But Larscheid says that isn’t the case and even compares the project with a fiasco in Transport Minister Andreas Scheuer’s ministry. “The app is Jens Spahn’s equivalent of (Scheuer’s) truck toll disaster – it cost a lot of money and has no apparent benefit.” The disaster is a reference to a truck toll technology system that Germany bought but failed to implement, at a cost of hundreds of millions of taxpayer money. “This app does more harm than good,” he says.
Larscheid believes the app would have to collect significantly more data to be useful – about the place and time of contact and also about the person, for example. “The app doesn’t even tell you if the alleged risk took place outdoors, on a commuter train or when visiting relatives in a hospital,” says Larscheid. The app’s algorithms operate with rough probabilities, he says, which isn’t effective enough. “You would never ride in an autonomous vehicle that might or might not hit a tree,” he says.
Other critics think the path chosen was the correct one, but that the execution has been less than stellar. “I have the impression that too many people are still only using the Warn-App out of their own interest – in other words, in the expectation of being warned themselves, but without the willingness to warn others in turn. But if too many people do that, the app can’t provide the full effect it was intended to,” says Anke Domscheit-Berg, the digital policy point person for the Left Party in parliament.
That refusal could explain one odd statistic. According to Deutsche Telekom, only 3,613 positive test results had been reported via the hotline as of last Tuesday. Even assuming that app users follow the hygiene rules and are less likely to belong to risk groups, that number would be concerningly small if you consider that there have been around 80,000 confirmed new coronavirus infections since the app’s launch.
It’s possible that many infected people aren’t reporting their positive test results to their app. Or that many of those who downloaded the app in the early days are no longer using it.
The diminishing interest in the app could also have something to do with the many glitches that have accompanied the project. At first, it didn’t update automatically in the background on some devices as intended, meaning the app didn’t warn people reliably for weeks in those instances. It also took quite a while for the people in charge to admit the mistake.
Now, a new glitch is annoying users. If you update your iPhone with the new operating system 13.7, the app might show an excessive risk of infection. The problem still hadn’t been fixed by the middle of last week.
Apple, Google and mobile phone manufacturers are partly to blame for some of these glitches. But the list of homemade bugs is also long. For example, only 134 of the 171 laboratories processing coronavirus test results are connected to the digital system behind the app. People whose tests are processed by the unconnected labs still have to call a special hotline to report their results if they are positive for the coronavirus and they have to provide their phone number, meaning that the process is by not nearly as anonymous as originally promised.
In addition, those wanting to get a coronavirus test must fill out a form on paper. Early on, printers weren’t able to meet the demand from doctor’s offices and health authorities for copies of the forms. Now, though, a new problem has cropped up. People have to mark their consent on a form in order to receive their results quickly through the app. Many people apparently overlook the box that must be checked, so they don’t get their results on their smartphone but, in many cases, through the mail. Losing time like that is a problem, especially when the goal is to quickly break infection chains.
“A Work in Progress”
Finally, there’s a study from Trinity College in Ireland that casts broader doubts about the efficacy of apps like the German one in buses, trains or planes. The report warns that the distance measurements taken by smartphones using Bluetooth are unreliable in those spaces. But the app was designed to be used in precisely those types of areas in Germany, because in the event of an infection, a person can quickly name and inform family members, friends or colleagues. But that’s not the case for people who have sat near that person on a bus or a commuter train. One of Germany’s Fraunhofer institutes for scientific research is now planning to carry out additional measurements in a high-speed train and in a passenger jet.
Those responsible for the app have sought to downplay the problems. The app, they say, is more of a “work in progress” rather than a finished product, just like any other IT solution.
And indeed, some of the apps that have been released elsewhere are performing much worse than the German app. Or, as in China, they are being used as nightmarish instruments of surveillance and control. And compared to other agonizingly slow digitization projects in Germany, the app was brought to market surprisingly quickly.
Still, the app doesn’t meet the government’s own expectations for it, which has left some disappointed. “In the beginning, there was a lot of hype about the app,” says Maria Klein-Schmeink, the health policy coordinator for the parliamentary group of the Green party. “The expectations were completely exaggerated.” Meanwhile, conservative Bavarian Governor Markus Söder has said that the app “isn’t a great success yet.” He added that considerations should be made for how to improve it.
Some have already made such considerations. SPD health expert Lauterbach and fellow party member Henning Tillmann, the co-chair of the digital policy organization D64, have offered their own suggestions for expanding the app’s utility. They suggest adding a voluntary contact diary so that people can log the people they have met with each day so that everyone can be warned in the event of an infection, including those who don’t use the app.
Software developer Tillmann says that newer findings about the spread of the virus should also be factored in. It could, for example, issue warnings in the event of particularly dangerous crowds of people, which the app can detect in real-time.
After around 100 days, though, this much is clear: The hope of having an app that can effectively help to contain the pandemic while at the same time offering comprehensive protections for private data has not been fulfilled – not in Germany and not in any other country in the world.
Health Minister Spahn has been pointing out for weeks that the app has been downloaded more often in Germany than in all other EU countries combined. Behind the scenes, experts in his ministry are already working on improvements. One idea they are looking into is whether people who don’t have modern smartphones could use Bluetooth wristbands in order to increase the number of app users. But that would be difficult to integrate into the current system.
More likely is that the claim will have to be abandoned that the app is some kind of super weapon in the fight against the pandemic. FDP politician Höferlin expresses those lower expectations as follows: “My attitude is: The app is voluntary, doesn’t collect much data, doesn’t cost users money or extra battery power, and if I only help to break a single chain of infection, then that’s already worth it.”