Here is a spreadsheet containing high-level Chinese policy documents (中发， 中办发， 国发， and 国办发) from the beginning of the first Hu administration (November 2002) through May 2016. I hope to update them semi-regularly. Please see the notes at bottom for more information about how we considered sourcing when choosing which documents to include.
I compiled these with the help of my friends and colleagues at MERICS – Julia Breuer, Max-Sebastian Dovi, and Matthias Stepan.
This information can also be found on my website at http://jessicabatke.com/policydocuments
To compile these data, we took advantage of the PRC’s very logical official document numbering system, and from February 15 to May 5, 2016, we conducted internet searches for these documents in sequential order. We searched for zhongfa (中发), zhongbanfa (中办发), guofa (国发), and guobanfa (国办发) issued from the beginning of Hu Jintao’s tenure as General Secretary (November 15, 2002) through the current year. For documents that were not listed in central government databases, we only included them in our data set if we could confirm their existence through two additional reliable sources (including sites such as local government websites, Xinhua articles, the Communist Party website, the Peking University law database, and Chinalawinfo.com). There are some documents which are released in full-text form but which do not themselves contain their document ID number. For these documents, we relied on separate sources that confirmed a match between the title of the document and the document ID number. Because of the documents’ sequential numbering identification system, we could be reasonably sure that certain documents do indeed exist, even if they are not publicly available or acknowledged. If we were unable to find the text or even the title of a document that should exist within the document numbering sequence, we counted it as “unknown.” For some Party documents, the full text is not released, but a rather detailed summary is, in which case we coded it as “summary.”
Of course, there are several issues with this collection method that might affect the accuracy of our data set. First, we cannot be completely certain how many documents of each type were released every year – there could be one (or several) “unknown” documents released at the end of any given year which we do not have a way to account for. This is because our collection method relies on sequential ordering and depends on having a “known” document follow an “unknown” document to ensure that the “unknown” document indeed exists. To try and offset this issue, we searched for ten additional documents (beyond the last “known” document) at the end of every year. It is unlikely that we are missing tens of “unknown” documents at the end of every year, but it is possible that we are missing a few in any given year. It is also possible that there are several documents that are indeed publicly available but that we were unable to find; that we did not count several apparently extant documents because we could not confirm their content through reliable sources; or that additional older documents have been released since we stopped compiling the data in May 2016.