## Reported Social Unrest Index: March 2022 Update

## Source details

**Canonical URL:** [Reported Social Unrest Index: March 2022 Update](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022084-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022084-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022084-print-pdf.pdf.json)

---

### Abstract and key findings
- This paper updates the Reported Social Unrest Index (RSUI) of Barrett et al. (2020), extending the dataset through to February 2022.
- Main findings:
  - Unrest was elevated during late 2019, coincident with widespread protests in Latin America.
  - Unrest fell markedly during the early stages of the COVID-19 pandemic as voluntary and involuntary social distancing increased.
  - Social unrest has since returned but generally remains below levels seen in 2019.
- JEL Classification Numbers: C82, F50, H80
- Keywords: Social unrest, Media coverage
- Author: Philip Barrett; Authorized for distribution by Malhar Nabar; May 2022; WP/22/84

### Methodology and background on the RSUI
- The RSUI proxies for major social unrest events by counting the number of media articles containing terms related to social unrest across a variety of sources.
- The measure is constructed consistently at a monthly frequency from the mid-1980s to June 2019 in Barrett et al. (2020); this paper provides an update through February 2022.
- Validation and robustness:
  - Large spikes in the index align closely with key events in authoritative narrative accounts.
  - The RSUI is robust to specifics of search criteria, sources used, and language of sources.
  - The RSUI performs at least as well as the Armed Conflict Location and Event Database for the tasks evaluated in Barrett et al. (2020).
- Uses of the RSUI and events:
  - Hadzi-Vaskov et al. (2021) and Hlatshwayo and Redl (2022) show real economic outcomes deteriorate following an unrest event identified by RSUI.
  - Hlatshwayo and Redl (2021) apply machine learning to predict unrest using RSUI-identified events as outcomes.
  - Barrett et al. (2021) use a daily version of the RSUI within event months to identify stock market responses to unrest events.
- Note: Events are the preferred cross-country comparator due to differences in media coverage and perception across countries.

### Trends in social unrest since July 2019
- The paper splits recent developments into three periods: pre-pandemic (second half of 2019), early pandemic (2020), and later stages of the pandemic (January 2021–February 2022).

- Pre-pandemic (second half of 2019):
  - Unrest was rising globally prior to the pandemic, a continuation of a trend since 2016.
  - A wave of protest in October and November 2019 concentrated activity in parts of Latin America, beginning in Chile and spreading more widely, causing sharp RSUI rises in several countries.
  - Unrest also rose in the Middle East with spikes in Algeria, Iran, Iraq, and Lebanon.

- Early stages of the pandemic (2020):
  - Unrest declined suddenly at the start of the pandemic, with the timing of decline coinciding closely with increased social distancing (both voluntary and mandatory).
  - Significant unrest events nonetheless occurred through the Summer of 2020, including:
    - Brazil: anti-government sentiment
    - United States: large protests for racial justice
    - Ethiopia: heightened inter-ethnic tensions
    - Lebanon and Belarus: large anti-government protests

- Later stages of the pandemic (January 2021–February 2022):
  - Social unrest continued in both Advanced Economies and Emerging and Developing Economies.
  - Advanced Economies examples where protests erupted (often anti-government or anti-lockdown): Canada, New Zealand, Austria, the Netherlands.
  - Emerging and Developing Economies examples and major events: anti-government protests in Kazakhstan and Chad; a coup in Burkina Faso; regional protests in Tajikistan; a constitutional crisis in Sudan.
  - Overall, social unrest remained low relative to pre-pandemic levels.

### Empirical illustrations and indicators referenced
- Figure summaries in the text (figures not reproduced here):
  - Figure 1: Fraction of countries with RSUI-identified events, 2016-2022 (monthly and three-month rolling average).
  - Figure 2a and 2b: RSUI January 2019 - January 2020, select South American and Middle Eastern countries (avg. = 100).
  - Figure 3: Unrest and mobility, 2018-2022 — comparing fraction of countries with social unrest events (3-month rolling average, LHS) and Google mobility data for transit stations (weekly average, RHS).
  - Figure 4: RSUI January 2020 - January 2021 for select countries (avg. = 100).
  - Figure 5a and 5b: RSUI January 2021 - February 2022 for select Advanced Economies and Emerging and Developing Market Economies (avg. = 100).

### Research implications and uses
- The updated RSUI dataset:
  - Enables monitoring of social unrest trends through February 2022.
  - Serves as an input for analysis linking unrest events to macroeconomic outcomes, financial market responses, and predictive modeling of unrest risk.
- Prior literature leveraging RSUI demonstrates its utility for IMF surveillance and assessments of social unrest vulnerabilities.

*Prepared by Philip Barrett; IMF Working Paper No. WP/22/84; May 2022.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022084-print-pdf.pdf_