A lot of attention worldwide is currently intensely focused on China's statistics; I'm not sure what gain there would be to their (or anyone's) willful manipulation of those. And, if only from a Chinese state perspective, I can see incentives in both directions - making the statistics look good as well as making them look bad. Either way, however, the statistics I used are from the Penn World Tables - and are the same data used by Pritchett and Summers (and Eichengreen et al, and many, many others). If I were to apply completely different data to overturn the PS reasoning, there'd always be the worry that the difference in opinion is due merely to different data. This way, it is clear the difference comes down instead to reasoning, rather than merely data.