In a highly interconnected world, pressures on one fragile state can have serious repercussions not only for that state and its people, but also for its neighbors and other states halfway across the globe. Since the end of the Cold War, a number of states have erupted into mass violence stemming from internal conflict. Some of these crises emerge from ethnic tensions; some are civil wars; others take on the form of revolutions; and many result in complex humanitarian emergencies.

Fault lines can emerge between identity groups, defined by language, religion, race, ethnicity, nationality, class, caste, clan or area of origin. Tensions can deteriorate into conflict through a variety of circumstances, such as competition over resources, predatory or fractured leadership, corruption, or unresolved group grievances. The reasons for state fragility are complex but not unpredictable. It is critically important that the international community understand and closely monitor the conditions that contribute to fragility — and be prepared to take the necessary actions to deal with the underlying issues or otherwise mitigate the negative effects.

To have meaningful early warning, and effective policy responses, assessments must go beyond specialized area knowledge, narrative case studies and anecdotal evidence to identify and grasp broad social trends. A mixed approach integrating qualitative and quantitative data sources is needed to establish patterns and trends. With the right data and analysis it is possible to identify problems that may be simmering below the surface. Decision makers need access to this kind of information to implement effective policies.

The Fragile States Index (FSI) produced by The Fund for Peace (FFP), is a critical tool in highlighting not only the normal pressures that all states experience, but also in identifying when those pressures are outweighing a states’ capacity to manage those pressures. By highlighting pertinent vulnerabilities which contribute to the risk of state fragility, the Index — and the social science framework and the data analysis tools upon which it is built — makes political risk assessment and early warning of conflict accessible to policy-makers and the public at large.

The strength of the FSI is its ability to distill millions of pieces of information into a form that is relevant as well as easily digestible and informative. Daily, FFP collects thousands of reports and information from around the world, detailing the existing social, economic and political pressures faced by each of the 178 countries that we analyze.

ORIGINS OF THE FRAGILE STATES INDEX: THE CAST FRAMEWORK

The genesis of most indices is to begin with a concept of what needs to be measured, followed by the development of a methodology that hopes to perform that measurement. The FSI followed a very different trajectory, whereby the idea for the Index occurred subsequently to the development of its own methodology.

FSI traces its origins to the creation of FFP’s Conflict Assessment System Tool (CAST), that was developed in the 1990s as a framework for policymakers and field practitioners to be able to better understand and measure conflict drivers and dynamics in complex environments. The CAST framework has been widely peer reviewed, and the continued usage of the framework by many of those same professionals, as well as now by local civil society and community groups in conflict-affected areas, is testament to the framework’s enduring relevance. In 2004, the CAST framework was used as the basis for the FSI, as researchers wished to determine whether state fragility could be assessed and ranked at a national level using the existing framework.

PRACTICAL APPLICATION: THE FRAGILE STATES INDEX ANALYTICAL PROCESS

Though at the ground level, the CAST framework is applied using various practices such as individual incident reporting and observation by field monitors, the sheer volume of data to be analyzed at an international level required a different approach. To that end, technology was employed to enable researchers to process large volumes of data to perform the national level assessments that feed in to the FSI.

Based on CAST’s comprehensive social science approach, data from three main streams — pre-existing quantitative data sets, content analysis, and qualitative expert analysis — is triangulated and subjected to critical review to obtain final scores for the Index.

  1. Content Analysis: Each of the twelve indicators of the CAST framework are broken down into sub-indicators, and for each of these, hundreds of Boolean search phrases are applied to global media data to determine the level of saliency of issues for each of those sub-indicators in each country. The raw data, provided by a commercial content aggregator, includes media articles, research reports, and other qualitative data points collected from over 10,000 different English-language sources around the world. Every year, the number of articles and reports analyzed is between 45-50 million. Based on the assessed saliency for each of the sub-indicators, provisional scores are apportioned for each country.
  2. Quantitative Data: Pre-existing quantitative data sets, generally from international and multilateral statistical agencies (such as the United Nations, World Bank, World Health Organization) are identified for their ability to statistically represent key aspects of the indicators. The raw data sets are normalized and scaled for comparative analysis. The trends identified in the quantitative analysis for each country are then compared with the provisional scores from the Content Analysis phase. Depending on the degree to which the Content Analysis and the Quantitative Data agree, the provisional scores are confirmed, or where they disagree, are reconciled based on a set of rules that dictate allowable movements in score in the event of disagreement between the two data streams.
  3. Qualitative Review: Separately, a team of social science researchers independently reviews each of the 178 countries, providing assessments based on key events from that year, compared to the previous one. Recognizing that every data set and approach has different strengths and weaknesses, this step helps to ensure that dynamic year-on-year trends across different indicators are picked up – which may not be evident in lagging quantitative data sets that measure longer term structural factors. It also helps to mitigate any potential false positives or negative that may emerge from noisy content analysis data.

These three data streams are then triangulated, applying a set of rules to ensure the data sets are integrated in a way that leverages the strengths of the different approaches. This approach also helps to ensure that inherent weaknesses, gaps, or biases in one source is checked by the others. Though the basic data underpinning of the Index is already freely and widely available electronically, the strength of the analysis is in the methodological rigor and the systematic integration of a wide range of data sources. Final indicator scores for each country are then produced based on from this process. A panel review is then conducted by the research team of the final index to ensure all scores are proportionate across the country spectrum.

The final FSI Index product is intended as an entry point into deeper interpretive analysis for the user – in line with the CAST framework approach. Though an index inherently ranks different countries – making some more fragile than others – ultimately the goal of the FSI is to measure trends in pressures within each individual state. By identifying the most salient pressures within a country, it creates the opportunity for deeper analysis and planning by policy makers and practitioners alike to strengthen each state’s resiliency. To that end, the following section outlines what each indicator seeks to measure in the Index – as well as providing guiding questions for deeper levels of analysis and inquiry by the user.