Privacy Valuation and Privacy-Seeking Behavior: Lessons from an Information Treatment

While information provision can update beliefs, it remains unclear whether belief-induced changes in valuation translate into concrete privacy-seeking intentions. To address this gap, we piloted a randomized information treatment with an instrumental variable strategy to causally identify the effect of post-treatment data valuation on intended privacy-seeking behavior (e.g., intentions to reduce data granularity and check privacy settings).

Respondents first report their beliefs about how accurately smart meter data reveal daily routines and state their valuation of hourly consumption data. The treatment group then receives factual information that predictive accuracy exceeds 90% before reassessing their beliefs and valuation. To identify the causal effect of valuation on behavior, we instrument post-treatment valuation with the interaction between pre-treatment belief error and the treatment indicator in a two-stage least squares framework.

Instrumental variable estimates show that higher privacy valuation causally increases intended privacy-seeking behavior. A one-euro increase in post-treatment valuation raises the probability of intending to provide less granular data by 0.8 percentage points and checking privacy settings by 0.7 points. These findings demonstrate that privacy-seeking behavior is valuation-driven and responsive to corrected misperceptions, with implications for data governance and digital energy policy.