 I'm Salvatore Bobonas and today's lecture is using data from China's National Bureau of Statistics. China has one of the most comprehensive English-language statistical websites of any non-English-speaking country. It's light years ahead of all other developing countries in this regard. China's statistical website may not be as good as the US Census Bureau or Eurostat for the European Union, but compared to other developing countries like Brazil, Mexico, Russia and India, it is fantastic. Conveniently, most Chinese data are broken down either by year or by province for easy comparison, and though there are difficulties comparing editions of the National Bureau of Statistics yearbook from year to year, you can generally find what you want if you look hard enough. There are many criticisms of the reliability of Chinese data, but most of it is probably reasonably accurate. That is most of the data, well, as well as most of the criticisms. Except for a few sensitive headline series like annual GDP growth, I think we can largely trust the numbers. In particular, demographic data from the MBS seem quite sound and seem to tell a meaningful story. The National Bureau of Statistics is China's main statistical agency and by far the most comprehensive English-language source for Chinese statistics. Most other sources you find on the web are simply repackages of data from the National Bureau of Statistics. The MBS tables are formatted as if they were tables in annual books. It's a bit old school, it's like getting a book and looking at a table instead of querying a database, but it's perfectly functional. The topics covered do differ year by year, which can make it difficult to navigate MBS data if you want to compare data series over multiple years. But for most purposes, it makes sense just to use the current year's data book, and if you need historical data for years that aren't currently reported, then go back to previous years and try to pull together the data you need. To find MBS data, just go to the web. If you Google National Bureau of Statistics of China, you'll easily get to the website. And most of the data that non-specialists will use are contained in the annual data series. You can either click on data or go to annual data. They both lead you to the same place. The most recent year for data is 2014, as of the time of this recording, which is March 2016. You can see that the data are broken up by topic. There are more than a dozen topics covered. In particular, population and national account data are covered in quite a bit of detail. I'm going to focus on the demographic data population, and first I'll show that the data, even politically sensitive data, are really forthcoming in these. For example, the floating population. The floating population in China is composed of people who have irregular hook-o status, usually people who have the right to live in a rural area, but who actually live in an urban area. You can see that there are now approximately 245 million people who are in this status. Something like a fifth of China's population have irregular status on our floating population. China distinguishes between those who just have a technical difference of residence and registration mismatch, and those that consider a floating population. But it doesn't describe exactly how it determines who is just a mistake. 289 million versus 245 million of those are a floating population. Note that China does tend to use units of 100 millions instead of units of millions or billions. So 2.45 here is actually 2.45 million, or 245 million. The same will be true of economic statistics, which are usually reported in hundreds of millions of yuan, and you have to make the appropriate comparison to get that to usually billions of yuan, will be what you're interested in. I'm going to scroll down here to table 211, which is an example of data by province instead of by year. These are dependency ratios. China doesn't always use the same definitions as other countries. China's dependency ratio is not really a ratio, it's actually a percentage. So 35% of the national population are either children or elderly. It doesn't define children and elderly. I believe that children are aged 0 through 16 and elderly are aged 64 and over, but don't cite me on that because I don't have a citation myself for that. But you can see that the child dependency ratio differs dramatically by province. The province with one of the highest is Guizhou province in southwestern China. 32% of the population are children. Compare that with Shanghai on the east coast, 11.7% of the population are children. And that illustrates the floating population problem right there. Adults from Guizhou go to work in Shanghai, leaving their children behind in Guizhou. You also see when you look at these provincial breakdowns that the provincial level units include both the provincial level cities like Beijing and Shanghai. They include provinces like Hebei and Shanxi, and they also include special regions like Inner Mongolia and Tibet. All of these are lumped together as provincial level units in China. Conveniently, the National Bureau of Statistics makes it very easy to download these data. You just press on the little icon here to download in Excel, and you get the file as an Excel file downloaded to your desktop or wherever you have downloads. The Excel file is formatted exactly the same as the view online. The data are grouped by provincial grouping, so these are the provinces around Beijing, then the northeastern provinces, then the east-central provinces, etc. And remember if you want to do any database operations that there is the national total row, and then there are blank rows separating each of the sets of provinces by row. Key takeaways. First, China's National Bureau of Statistics is a highly professional statistical agency by developing country standards. It may not stand up to the US Census Bureau or the European Union's Eurostat, but compared to any other developing country, it's a fantastic outfit. Second, economic and demographic data have particularly good coverage in the NBS annual data book. Third, often data for earlier years are actually available in earlier editions, but the table numbers change from year to year, so it can take a little bit of work to go back and reconstruct data series for multiple years. Thank you for watching this video. For more about me, go to Salvatourbebonus.com, where you can also register to receive my monthly newsletter on global current affairs.