Skip to main content
. 2021 Nov 25;151:105753. doi: 10.1016/j.worlddev.2021.105753

Fig. 2.

Fig. 2

Comparison of Covid-19 (mid-2020) and HIV (2006) Perceptions Panel A: We combine data on the perceived prevalence and perceived mortality to create a measure of excess mortality. Perceived excess mortality for HIV was estimated using the difference between the probability of a hypothetical person dying in 5 years if sick with HIV/AIDS minus the probability of a healthy person dying in 5 years. Perceived excess mortality for Covid-19 was estimated as the probability of dying from Covid-19 conditional on being infected. Panel B: Likelihood of infection with Covid-19 (2020) and HIV (2006) is obtained using a method to elicit subjective probabilistic expectations that has been implemented in the MLSFH since 2006 (Delavande and Kohler, 2009). In the 2020 MLSFH Covid-19 Phone Survey, this question was worded as “Out of 10, tell me the number of peanuts that reflects how likely you think it is that you are infected with coronavirus (Covid-19) now?,” where each peanut represents a 10% chance. An analogous question was asked about the likelihood of being infected with HIV in 2006. “In the future” refers to 3 months for Covid-19 and 2 years for HIV in 2006.