[Berobero]

There's definitely reason to be suspect, although it looks like, on the average, it might be the reverse of what you seem to be expecting (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2785092/):

"Our model suggests that persons sleeping 5 and 7 hours over-reported, on average, by 1.3 and 0.3 hours respectively."

For self-reporting sleep studies, perhaps people have it in their head that 8 hours is how much they're "supposed" to sleep and they over-report to better match that?

Of note, however, while the CDC data I linked comes from the Behavioral Risk Factor Surveillance System which does appear to be merely a self-reporting survey question (https://www.cdc.gov/sleep/surveillance.html), the data in the Economist article comes from an app (www.sleepcycle.com). Granted, I suspect the latter almost surely suffers from some kind of non-trivial selection bias.