What was the effect of Israel’s decision to use mass surveillance of cellular phones for COVID-19 contact tracing? Oshrat Ayalon and I have uploaded a new manuscript to arxiv that analyzes the relationship to the installation of the more accurate app.

Israel rolled a contact tracing app at the same time as employing “The Tool”, a cellular tracking system by secret service, for COVID. We use this as a natural experiment to see how attitudes toward mass surveillance affect people’s installation of the contact tracing app. The paper analyzes a survey carried out in May 2020 in Israel, with a representative sample of 519 participants (Heb + Arb speaking). We have asked whether people have installed or uninstall the app and asked about attitudes towards both systems.

We found that positive attitudes toward The Tool were related to a reduced likelihood of installing the app and an increased likelihood of uninstalling them. This was strong even after controlling for privacy perceptions, the utility of systems, demographics, COVID concerns, trust in gov, etc.

We think that The Tool “crowded out” voluntary use of the app, which was tragic: the app was more accurate but the dwindling installation base (after a promising start) had rendered it useless. We hope that our study can serve as a warning against future mistakes.

We have also released the raw data (after removing identifying and pseudo-identifying properties) and the code of our analysis. If you are interested in analyzing contact tracing apps behaviors and attitudes, you are more than welcome. Code and data are on Git.