Website: www.policychangeindex.com
Authors: Julian TszKin Chan and Weifeng Zhong
Please email all comments/questions to ctszkin [AT] gmail.com or weifeng [AT] weifengzhong.com
China's industrialization process has long been a product of government direction, be it coercive central planning or ambitious industrial policy. For the first time in the literature, we develop a quantitative indicator of China's policy priorities over a long period of time, which we call the Policy Change Index (PCI) of China. The PCI is a leading indicator that runs from 1951 to the most recent quarter and can be updated in the future. In other words, the PCI not only helps us understand the past of China's industrialization but also allows us to make short-term predictions about its future directions.
The design of the PCI has two building blocks: (1) it takes as input data the full text of the People's Daily --- the official newspaper of the Communist Party of China --- since it was founded in 1946; (2) it employs a set of machine learning techniques to "read" the articles and detect changes in the way the newspaper prioritizes policy issues.
The source of the PCI's predictive power rests on the fact that the People's Daily is at the nerve center of the China's propaganda system and that propaganda changes often precede policy changes. Before the great transformation from the central planning under Mao to the economic reform program after Mao, for example, considerable efforts were made by the Chinese government to promote the idea of reform, move public opinion, and mobilize resources toward the new agenda. Therefore, by detecting (real-time) changes in propaganda, the PCI is, effectively, predicting (future) changes in policy.
For details about the methodology and findings of this project, please see the following research paper:
- Chan, Julian TszKin and Weifeng Zhong. 2018. "Reading China: Predicting Policy Change with Machine Learning." AEI Economics Working Paper No. 2018-11 (latest version available here).
Results will change as the underlying models improve. A fundamental reason for adopting open source methods in this project is so that people from all backgrounds can contribute to the models that our society uses to assess and predict changes in public policy; when community-contributed improvements are incorporated, the model will produce better results.
Three python functions and an R script are available for users to process data and build the neural network models to construct the PCI of China.
proc_pd.py: Process and prepare the raw data from the People's Daily for building the neural network models.pci.py: Train a neural network model to construct the PCI for a specified year-quarter.compile_model_results.py: Compile the results from all models and export them to a.csvfile.generate_figures.r: Generate figures.
Users can check out the descriptions of the arguments for each python function using the --help option. For example:
./PCI-China/data>python proc_pd.py --help
Using TensorFlow backend.
usage: proc_pd.py [-h] [--input INPUT] [--create CREATE] [--seed SEED]
[--k_fold K_FOLD] [--output OUTPUT]
optional arguments:
-h, --help show this help message and exit
--input INPUT Input file
--create CREATE 1:Create database 0:Append
--seed SEED Seed
--k_fold K_FOLD k_fold
--output OUTPUT Output filename
./PCI-China>python pci.py --help
Using TensorFlow backend.
usage: pci.py [-h] [--model MODEL] [--year YEAR] [--month MONTH] [--gpu GPU]
[--iterator ITERATOR] [--root ROOT] [--temperature TEMPERATURE]
[--discount DISCOUNT] [--bandwidth BANDWIDTH]
optional arguments:
-h, --help show this help message and exit
--model MODEL Model name: window_5_years_quarterly,
window_10_years_quarterly, window_2_years_quarterly
--year YEAR Target year
--month MONTH Target month
--gpu GPU Which gpu to use
--iterator ITERATOR Iterator in simulated annealing
--root ROOT Root directory
--temperature TEMPERATURE
Temperature in simulated annealing
--discount DISCOUNT Discount factor in simulated annealing
--bandwidth BANDWIDTH
Bandwidth in simulated annealing```
```{shell}
./PCI-China>python compile_model_results.py --help
usage: compile_model_results.py [-h] [--model MODEL] [--root ROOT]
optional arguments:
-h, --help show this help message and exit
--model MODEL Model name: window_5_years_quarterly,
window_10_years_quarterly
--root ROOT Root folder
The raw data of the People's Daily, which are not provided in this repository, should be placed in the sub-folder PCI-China/Input/pd/. Each file in this sub-folder should contain one year-quarter of data, be named by the respective year-quarter, and be in the .pkl format. For example, the raw data for the first quarter of 2018 should be in the file 2018_Q1.pkl. Below is the list of column names and types of each raw data file:
>>> df1 = pd.read_pickle("./PCI-China/Input/pd/pd_1946_1975.pkl")
>>> df1.dtypes
date datetime64[ns]
year int64
month int64
day int64
page int64
title object
body object
id int64
dtype: object
where title and body are the Chinese texts of the title and body of each article.
The processed data of the People's Daily, which are not provided in this repository, should be placed in the sub-folder PCI-China/data/proc/by_year/. Each file in this sub-folder should contain one year-quarter of data, be named by the respective year-quarter, and be in the .pkl format. For example, the processed data for the first quarter of 2018 should be in the file 2018_Q1.pkl. Provided with the raw data, the processed data are the output of running proc_data.sh. Below is the list of column names and types of each processed data file:
>>> df2 = pd.read_pickle("./PCI-China/data/proc/by_year/2018_Q1.pkl")
>>> df2.dtypes
date datetime64[ns]
year int64
month int64
day int64
page int64
id int64
quarter int64
weekday int64
frontpage int32
page1to3 int32
title_len int64
body_len int64
n_articles_that_day int64
n_pages_that_day int64
n_frontpage_articles_that_day int32
training_group int32
title_int object
body_int object
dtype: object
where title_int and body_int are the word embeddings (numeric vectors) of the title and body of each article.
Please cite the source of the latest Policy Change Index (PCI) of China by the website: http://www.policychangeindex.com.
For academic work, please cite the following research paper:
- Chan, Julian TszKin and Weifeng Zhong. 2018. "Reading China: Predicting Policy Change with Machine Learning." AEI Economics Working Paper No. 2018-11 (latest version available here).