2018 passed without any regional waves of social unrest akin to the Arab Spring or the “color” revolutions of the early 2000s, but several countries still saw significant bursts of protest activity in the past year, and some of those bursts produced major political change.

We can track protest activity around the world in real time and in retrospect with Kensho’s topic trend scores. These scores are a weighted, normalized count of news articles associated with a specific country and tagged with a particular topic. The topic tags used in these computations are generated by an array of binary classifiers, each trained on hundreds of hand-labeled examples. The weights in the trend-score tallies represent the prominence of the country of interest in each article’s text, as determined by a separate machine-learning process. To obtain the final score, these weights are grouped by day, summed, and then divided by the total count of articles on all topics. The end result is a time series that tracks the amount of “chatter” about a given country-topic pair in a particular set of news sources.

Kensho produces topic trend scores for all countries worldwide every day on dozens of geopolitical themes. The “protest” tag is designed to spot articles in which “gatherings in public spaces for the purpose of making political or economic demands or claims, including but not limited to demonstrations, pickets, sit-ins, strikes, walkouts, general strikes, and riots” represent a major theme. These articles don’t have to describe specific protest events to warrant the tag, but they often do, and fluctuations in trend scores for this topic generally do a solid job tracking variation over time in protest activity within and across countries.

So, what do our protest trend scores tell us about where social unrest kicked up around the world in 2018?

When the year started, Iran was already in the throes of one of the biggest and broadest waves of protest activity it has seen since the disputed elections of 2009, and perhaps since the Islamist revolution. These protests erupted in response to frustration over rising prices and taxes, and they quickly grew to include an unusually large number of cities and diverse array of participants. As the chart below shows, those demonstrations ebbed early in 2018, but unrest persisted throughout the year as Iran slid into an economic crisis exacerbated by the tightening of U.S. sanctions.


While the unrest in Iran has not (yet) produced major political change at the national level, popular uprisings in Ethiopia and Armenia did. In Ethiopia, ethnic and regional frustrations that spawned a wave of protests in 2016 flared anew in early 2018, causing a new flurry of demonstrations that led to the surprise resignation of the country’s prime minister in mid-February. His replacement has pursued reforms that have spurred talk of a “democratic awakening” in Ethiopia, and that probably contributed to an attempted assassination at a rally in June.


A popular uprising also toppled an authoritarian leader and launched a process of democratization in Armenia. When longtime president Serzh Sargsyan tried to ram through constitutional reforms that would have extended his hold on power while shifting him to the office of prime minister, tens of thousands of Armenians took to the streets and eventually won the backing of the country’s security services. Sargsyan quickly resigned and, as anticipated on this blog, his replacement has moved the country’s politics in a more democratic direction.


One of the biggest surges in protest activity in 2018 occurred in Nicaragua, where demonstrations over social-security reform in April morphed into a sustained rebellion against Daniel Ortega’s authoritarian rule. As the chart below implies, protest activity waned as the year passed and state repression intensified, but the confrontation persisted into the new year.


Two countries that saw surges in protest activity in late 2018 were France and Sudan. In France, protests called by the “yellow vest” movement swept over the country in November and December after months of organizing around a basket of social concerns that included an unpopular fuel-tax hike. These protests proved unusually persistent and disruptive, and they drove President Macron to delay the fuel-tax increase, to promise to raise the minimum wage, and to organize “town hall” meetings at which citizens could air grievances with the government.


Rising costs of living also played an important role in sparking fresh unrest late last year in Sudan. Austerity measures imposed in mid-December spurred street protests in numerous towns. Those protests rapidly spread and expanded their demands to include the overthrow of the incumbent regime, drawing support from the country’s repressed opposition parties. Many observers have compared the country’s current protest wave to uprisings that toppled governments in 1964 and 1985 (see here and here, for example), but few experts so far have confidently predicted the departure of President Bashir or his allies in parliament.


Finally, frustrations over unemployment and poor living conditions also spurred a wave of protests in Iraq’s southern provinces, in and around the city of Basra in particular. In early July, neighboring Iran cut electricity supplies to Iraq during a brutal heat wave and on top of a ongoing water shortage, and those conditions apparently tipped social unrest in the region from a simmer to rolling boil. As reflected the trend-score chart below, protest activity around Basra ebbed in August before flaring anew for a particularly intense week in early September, then settled into a steadier rhythm for the rest of the year.


These seven countries were not the only ones that saw significant protest activity in 2018, but they do represent some of the most notable cases. They also show how Kensho’s “protest” topic tag can be used to spot new waves of unrest as they emerge, and to help track the dynamics of ongoing ones.

Jay Ulfelder

Political science Ph.D. (Stanford 1997), research director for the US government-funded Political Instability Task Force (2001-2011), Good Judgment Project superforcaster.