Great ResignationClimate ChangeLeadershipInflationUkraine Invasion

Americans don’t trust contact tracing apps. Here’s how we can fix that

August 17, 2020, 4:31 PM UTC

With COVID-19 cases climbing across the U.S. and manual contact tracing efforts to stem the spread of the disease faltering, officials are turning to a tech solution that nearly four dozen countries have adopted: digital contact tracing apps.  

In the coming weeks, over 20 states and localities in the U.S. plan to roll out COVID-19 contact tracing apps based on a model developed by Google and Apple—one that relies on a phone’s Bluetooth signal to exchange digital handshakes with other devices in a six-foot range.

The data collected from the app is stored within a user’s phone, rather than on a government server, and only when an app user is diagnosed with COVID-19 are they asked by health authorities to share the data in the app so that others can be alerted to a possible infection. These apps become more effective at preventing the spread of COVID-19 if more people download and use them. Despite some countries’ botched rollouts, Ireland has shown that it’s possible to get a large percentage of the public to use these apps. 

But deploying contact tracing apps is an uphill battle in the U.S.: Americans are still too distrustful of contact tracing apps for them to be effective. That means public health authorities have a lot of work to do to allay the public’s fear before they can deploy this technology.

In a study of 2,000 Americans, fielded between June 25 and 26, 2020, we found that 42% supported the adoption of contact tracing apps. Support was lower for digital contact tracing than for five other public health surveillance policies: temperature checks, traditional contact tracing, centralized quarantine, immunity passes for public transit, and electronic device monitoring. 

In open-ended survey responses, many respondents expressed concerns about the government tracking their location or retaining their data for future use. “Another step toward socialism,” one wrote. “I don’t want the government…knowing every little thing I do.” Similarly, another reported that “I despise seeing our Constitution shredded to pieces by people who do not have our best at heart.” Others warned that “tracking people via phone would never end” and that the app would put “too much power in the wrong hands.”  

Consistent with this, we also found higher support for hypothetical contact tracing apps that incorporate privacy protections. We used conjoint analysis, a tool used in marketing research, to figure out which features make apps more appealing to the public. We find that the public is more willing to use apps, like those supported by the Google-Apple API, that store user data on phones rather than on a central server. Forty-four percent of respondents say they would download a hypothetical app that stores data on users’ phones, compared with 39% when the app was described as storing data on a central server. This is good news given that the Google-Apple API has emerged as the dominant model—in contrast to earlier efforts by governments in Asia and the U.K. that involved centralized data storage—among public health authorities rolling out digital contact tracing in the U.S.   

Although the Google-Apple API had been designed with privacy in mind, there are still fixes to be made before rolling out apps in the U.S. For instance, Google needs to ensure that enabling Bluetooth on Android phones does not automatically turn on location tracking. Public health authorities need to make sure that the public understands what the technology is doing. Even if an app is privacy-preserving, the public may not perceive it as such. After all, the public has reasons for their distrust: Apps that have been deployed so far in the U.S. track GPS location data. Even more shockingly, North Dakota’s contact tracing app (released prior to the Google-Apple API) had sent location data and unique user identifiers to Foursquare and other data to Google, despite promising not to send data to third parties. Such violations of user privacy will likely turn potential users away.   

Beyond privacy concerns, political polarization could impact adoption of contact tracing apps, just as wearing masks has become a partisan issue. In our survey, 47% of Democrats and 46% of Republicans support the government encouraging everyone to download and use contact tracing apps. But 39% of Republicans, compared with 27% of Democrats, oppose such a policy. These partisan divisions could increase over time as public awareness of digital contact tracing grows. Republican governors and elites have a window to stave this off by encouraging adoption and emphasizing privacy-preserving features.  

Finally, public health officials must recognize that many Americans, often those who are the most vulnerable to COVID-19, cannot use contact tracing apps. For instance, only 53% of those who are 65 and above own smartphones. Smartphone ownership is also lower among those with lower levels of education and income, who are also more likely to be working “essential” service jobs. To protect these populations, public health authorities should continue to invest in improving manual contact tracing. In our survey, we find that Black Americans are less supportive of expanding traditional contact tracing compared with white Americans. Considering these survey results, states should recruit and train contact tracers who can effectively communicate to those in minority communities who are overpoliced and may fear overreach by the government.  

The deployment of contact tracing apps need not be a disaster. Ireland’s app, using the Google-Apple API, has become a relative success. As of July 31, 37% of Ireland’s population over the age of 15 has downloaded the app. Given its success, which has been attributed to trust in government and a positive user experience, Ireland donated the code for its app to the Linux Foundation Public Health initiative so that other countries can build similar apps.   

The basis of Ireland’s success—trust in government and a positive user experience—highlights challenges facing adoption in the U.S., where trust in government is at an all-time low. Although technology can address the user experience, there are no easy tech fixes to trust in government. 

Fortunately, trust in local and state governments remains high. Upcoming contact tracing rollouts would be wise to emphasize the local and state basis for app implementation. Rolled out along these lines, the U.S. is more likely to find that the answer to its national public health crisis resides in a patchwork of local success stories.  

Sarah Kreps and Baobao Zhang are political scientists at Cornell University. 

Nina McMurry is a political scientist at the Massachusetts Institute of Technology.