Public-trust effects of transparent differential privacy protections

Determine the impact of publishing detailed differential-privacy protections, including the level of risk to which respondents are exposed, on public trust in National Statistical Organisations and on the perceptions of potentially affected subpopulations.

Background

Differential privacy permits National Statistical Organisations to publish the details of their privacy mechanisms and perturbation distributions without reducing the formal privacy guarantee. This transparency can help sophisticated analysts account for the added uncertainty in statistical outputs, but typical users may find the resulting variance analysis difficult to interpret.

The paper identifies an unresolved social and communication issue: transparency might reassure the public that an organisation is protecting confidential information, or it might alarm subpopulations by making their exposure to privacy risk more visible.

References

However, it is unclear what impact this transparency will have on public trust: it could reassure the public of NSOs’ safety or, by publishing how ‘at risk’ a respondent’s information is, it could scare certain subpopulations [16].

— Big data, differential privacy, and national statistical organisations  (2609.02495 - Bailie, 2 Sep 2026) in Section 3.4, “Advantages of DP”