An article on Caixin Online is an interesting example of how national and provincial statistics are "compiled" in China. Schoolhouse Shock exposes the dismal state of public education in poor rural areas. This is of course terrible for the long-run economic prospects of the residents (and the overall economy). But it also shows how all the educational statistics are completely fabricated. Regarding drop-out rates, the article says:
Some officials at schools in Gansu Province said on condition of anonymity that the dropout rate is roughly 30 to 40 percent in junior high schools.
Provincial government statistics put the dropout rate at 0.6 percent. Central government figures for Gansu state the dropout rate is lower than 3 percent and declining.
This should be a wake-up call to those tempted to give China's National Statistics Bureau the benefit of the doubt. If educational statistics are so far from reality, why would the economic ones be any better?