Skip to main content

Understanding disorder in Central Asia

Countries
Kazakhstan
+ 4 more
Sources
ACLED
Publication date
Origin
View original

In this joint report by the Armed Conflict Location & Event Data Project (ACLED) and the Oxus Society for Central Asian Affairs, Dr. Roudabeh Kishi, Asena Karacalti, Edward Lemon, and Bradley Jardine review new data on political violence and protest in Kazakhstan, Kyrgyzstan, Uzbekistan, Tajikistan, and Turkmenistan.

The five countries in Central Asia — Kazakhstan, Kyrgyzstan, Uzbekistan, Tajikistan, Turkmenistan — all display varying degrees of authoritarian rule. Across the region, local economies are stymied by corruption, resulting in a dissatisfied and increasingly vocal populace (Carnegie Endowment for International Peace, 18 October 2019). In two of the republics, Kazakhstan and Kyrgyzstan, election-related demonstrations have altered the political landscape, but across the board, state directed-force and mob brutality are rife.

This report builds on a burgeoning partnership between the Armed Conflict Location & Event Data Project (ACLED) and the Oxus Society’s new Central Asia Protest Tracker (CAPT). While the two datasets were designed to tackle different research questions, the data they provide and the trends that they highlight are complementary, allowing users to gain an even broader understanding of political dynamics in the region.

Introduction to ACLED

ACLED is a disaggregated data collection, analysis, and crisis mapping project tracking disorder, which includes political violence, demonstrations, and strategic developments. Information is tracked around the world in real-time, collecting information on the dates, actors, locations, fatalities, and types of events that take place. The ACLED project’s methodology covers a range of events types, including battles, explosions/remote violence, violence against civilians, protests, riots, and strategic developments. These event types are further disaggregated into sub-event types for a more granular analysis of disorder. In addition to information on actors involved in events, the type of agent is also captured to track broader shifts in patterns of disorder.

In addition to data collection, the ACLED team conducts analysis to describe, explore, and test conflict scenarios, with analysis made freely available to the public. ACLED’s analysis is developed from extensive academic research into the dynamics of political violence across the world and specific case studies of conflict agents, local trends, and intersections with domestic political contexts. ACLED analysis is unique due to its intersection between theoretically informed frameworks and hypotheses with local-level empirical data and analysis.

While ACLED is a United States-based non-governmental organization with a 501(c)(3) designation, it is a fully remote organization, with researchers based around the world. Such a system allows for researchers to be based in or have connections to the countries that they cover, benefitting from local language and context knowledge. ACLED data rely on a number of sources, including traditional news media, with local news sources prioritized; civil society organizations; local partners, such as local conflict observatories and organizations tracking certain political violence or protest trends in specific contexts; and select ‘new media’ sources. Given the restrictions on the free press in the region, trusted new media sources are used to expand coverage, especially for rural events where they help to capture local, small-scale protests and mob events. *ACLED does not crowdsource information. *New media relied upon in the Central Asian region includes broadcasting services reporting through videos posted to online forums, Facebook pages of local political groups, and Twitter accounts of prominent politicians and specialists of the region, for example. A targeted approach to the inclusion of new media in Central Asia has been adopted by ACLED, as all sources are verified for reliability and only reports with clear information on location, actor, and date are coded. This overarching sourcing strategy allows for the production of locally informed data, which are the cornerstone of ACLED’s methodology. This approach sets ACLED apart from other global conflict datasets in critical ways.

The ACLED dataset is hence built on the notion that all countries experience some forms of disorder, and that the modality of that disorder is strongly shaped by the specific vulnerabilities, political fault lines, and priorities of each society. Waning institutional trust and strength, income inequality, and demographics can all be sources of disorder, contributing to both demonstrations as well as political violence. Protest movements and government response to demonstrators can result in the birth of violent actors engaging in conflict, and ongoing conflicts can fuel protests involving affected populations, for example. This is why ACLED seeks to capture this spectrum of disorder, especially given that disorder does not present as the same in every space. An expansive view of its modalities allows for an accurate account of the occurrence and patterns of disorder across countries as diverse as those in Central Asia in comparison to those in Africa or Latin America, for example. ACLED hence relies upon a standardized methodology to capture the spectrum of disorder around the world, while weighing methodological decisions whenever expanding coverage to new geographic regions. (For example, specific methodological decisions taken into account for Central Asia are described in this primer.)

ACLED’s Central Asian coverage dates back to the beginning of 2018, and as of the end of 2020 includes 2,874 events capturing demonstrations, political violence, and strategic developments, including 1,589 in Kazakhstan, 814 in Kyrgyzstan, 336 in Uzbekistan, 97 in Tajikistan, and 38 in Turkmenistan. This information comes from over 130 unique sources. Data are updated weekly by experienced researchers who review traditional media in Kazakh, Tajik, Turkmen, Kyrgyz, Uzbek, Russian, and English languages.

Introduction to Oxus Society’s CAPT

The Central Asia Protest Tracker (CAPT) is an original dataset of protest events compiled by the Oxus Society for Central Asian Affairs. It describes and categorizes protest activity in each of the five Central Asian republics from 1 January 2018 to 31 December 2020 with future updates to be scheduled on a quarterly basis. Currently, the Central Asia Protest Tracker records 1,577 protest events, including 780 in Kazakhstan, 603 in Kyrgyzstan, 147 in Uzbekistan, 29 in Tajikistan, and 18 in Turkmenistan. The dataset provides a useful analytical tool for the study of local protest dynamics and the issues which prove to be of most importance to the peoples of the region. In addition to issue sets, the data also offer insight into common targets of dissent, whether it is local or national governments, businesses, or law enforcement — to name a few, as well as the responses made by targeted institutions, ranging from the use of force to accommodating protester demands.

The CAPT database consists of online published material that is public and identifiable, but not private. The use of online source material allows Oxus Society to cross-check data and conduct follow-up studies, for example, by reaching out to activists involved in specific protests to inquire about whether their activities led to substantial policy change. Due to the nature of the collected data, the dataset is likely to have a substantial effect on the results of empirical research. Reliance on reporting means that many cases of activism not considered “newsworthy,” or that took place in geographically remote communities, may be missing from the final results, skewing findings. While the CAPT does attempt to address this by mining social media platforms for additional sources of data, these sources also come with their own issues of verification and many unreported protests are likely missing from the dataset. The CAPT dataset uses ACLED entries to supplement its own data gathering by identifying issue cases; the research team searches corresponding reporting from the incidents to verify events and data according to CAPT coding standards.

The CAPT uses 28 categories of issue type ranging from environmental protests to pro-government demonstrations, while also coding for 9 types of target (e.g. national government, security services, foreign businesses). This degree of coding allows for a much more granular approach to understanding protest dynamics and the issues which are mobilizing Central Asians on the streets.

Both ACLED and CAPT rely on a geospatial definition of dissent, which leaves out important activities like online protests, petitions, and types of resistance that do not require citizens taking to the streets. This narrow definition of dissent therefore skews the results, with important forms of subtle dissent and online campaigns left out of the analysis.

Complementarity of the Data Projects

Given the data coverage and mandates of these projects, the picture of Central Asia that emerges based on data from the two is complementary, with patterns of overlap and distinction between the two. The maps below depict this overlap by matching events across the two datasets. In addition to the CAPT events which cite ACLED explicitly, there is also overlap in coverage across other events as well — those events which both datasets coded based off of similar sources, for example. In order to ensure an accurate reflection of the overlap, such events have also been matched. Events colored in navy are those which both organizations report. Events in blue are those which ACLED alone reports; this includes some protests, as well as incidents of political violence, including communal violence, state repression, and mob violence, which fall outside of CAPT’s threshold for inclusion. Events in orange are those which CAPT alone reports, which are largely small-scale acts of protest, like self immolations or individual-person demonstrations; these events fall outside of ACLED’s threshold for inclusion, which identifies protests as groups of three or more people physically congregating together.