The Policy Change Index (PCI) is a series of open-source machine learning projects that predict authoritarian regimes’ major policy moves by “reading” their propaganda publications. The first three projects inducted into the series are about China’s policies and based on its official newspaper — the People’s Daily.

  1. PCI-China: predicts China’s policy changes from 1951 Q1 to the present.

  2. PCI-Crackdown: predicts how close in time the 2019-20 Hong Kong protests are to a Tiananmen-like crackdown by China.

  3. PCI-Outbreak: measures the severity of an epidemic outbreak in China, such as COVID-19.

PCI-China and major events in China, 1951 Q1 to 2021 Q2

A spike in the PCI-China signals a major policy change, while a vertical bar marks the ground truth of the change labeled by the event. The PCI-China often spikes months before policy changes take place, validating the index’s predictive power. Click here to learn more about how it works.

PCI-Crackdown for 2019-20 and 2014 Hong Kong protests

The closer the PCI-Crackdown gets to the June 4 line, the higher the possibility of a Tiananmen-like crackdown. The PCI-Crackdown for the 2019-20 Hong Kong protests remained within three weeks from the crackdown line throughout 2019. The issue reemerged in 2020 as China pushed for a national security law. In comparison, the PCI-Crackdown for the 2014 Hong Kong protests is lower and downward-trending. Click here to learn more about how it works.

2019-20 Hong Kong protests

Second wave: Apr 16 to Jul 20, 2020

First wave: Jun 9, 2019 to Jan 7, 2020

2014 Hong Kong protests, Sep 26 to Dec 15

PCI-Outbreak for COVID-19 and official statistics in China (Jan 21 to Sep 15, 2020)

The index uses the 2003 severe acute respiratory syndrome (SARS) as the benchmark. The higher the indicator, the larger the scale of the outbreak. Click here to learn more about how it works.


Sign up to stay up to date on the series of PCI projects.


    What’s new


    What’s next