Disclosure in epidemics

We study information disclosure as a policy tool to minimize welfare losses in epidemics through mitigating healthcare congestion. We present a stylized model of a healthcare congestion game to show that congestion occurs when individuals expect the disease to be sufficiently severe and this leads to misallocation of scarce healthcare resources. Compared to full disclosure, under which congestion occurs when the true severity level surpasses the exhaustion level, a censorship policy, which pools the true severity levels around this exhaustion level and fully reveals all other severity levels, helps to reduce congestion and is welfare improving. Under mild conditions, we show that such a policy is indeed optimal. We further show that this insight is robust to considering partially effective pre-screening and limited information leakage.