Interestingly, you can get multiple-wave dynamics even when very far from population-level herd immunity, and even without taking into account changes in behavi

Interestingly, you can get multiple-wave dynamics even when very far from population-level herd immunity, and even without taking into account changes in behavior (for example, [1]). That's because of stochasticity in the process due to the fact that (1) most people infect no one, but some people infect many ("overdispersion") and (2) the population is not well mixed. In practice, of course, people respond to news about Covid waves and that contributes to the wave-like dynamics. Furthermore, new variants of the virus will be able to reach people that were "cautious enough" for a previous variant. That alone would also give you wave-like dynamics, basically every new variant would affect increasingly harder to reach people. Anyway, the point is, seeing these waves is super natural and expected and we can't deduce anything about population-level immunity from them without looking really really closely. [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4526984/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4526984/