If the negative correlation between deadliness and contagiousness is not real, then why did it ever exist? First, it makes sense. A microbe does not “care” if you die; its only goal is to hang around long enough to infect another host. From that perspective, it makes sense why a pathogen might evolve to become less deadly: A dead host is useless. So too is a really sick host (who is unable to go anywhere). Thus, the thinking goes, the optimal strategy is to make a host sick enough to spread the pathogen (for example, through coughing or vomiting) but not too sick. Herpes, the common cold, and a whole slew of other pathogens appear to have adopted this strategy.
Second, a compelling line of evidence came from the purposeful extermination of an invasive species of rabbit in Australia. Authorities released the myxoma virus into the rabbit population, which killed almost all of them. However, as time went on, rabbit mortality from the virus dropped from 99.5% to 90% — providing evidence for the “declining virulence” hypothesis. (A somewhat recent observation that some of the viruses have regained potency by suppressing the rabbits’ immune system adds a new wrinkle to the narrative.)