That’s not really my point.
You can’t cancel out bad things with good things.
That China averages out to stats that look OK, doesn’t make the stuff happening in HK or Xinjiang go away.
If you cut up the population groups into multiple population sized chunks with the compared countries, they each will still have a better number.
Not necessarily. If you average everything, that hides any outliers.
The bigger the sample size for your stats, the less it tells you. You could have an entire country’s worth of people in China, living in conditions that lead to horrible stats, and have that essentially disappear in the stats of the whole country.
You should break stats down into smaller chunks. But not just arbitrarily. Do it by province, city, district. That’s how you find the problem areas. The bigger the area you measure, the less the stats tell you about the individual citizens.
The increase is recent, and still rising.