An Introduction to the Business and Conflict Barometer
What if policymakers, investors, managers, and community advocates could easily ‘listen in’ on the dynamics of conflict and cooperation touching the private sector, within any defined timeframe and localized geographic area on the planet? This would enable the identification of conflict-prone business and investment practices, as well as of peace-positive pathways for business development. These would have practical importance for planning, monitoring, and learning, both at the project level and at global scale.
This is the promise of the Business and Conflict Barometer (BCB), the public-facing version of which is now under advanced development. Below, we will introduce you to the BCB, and illustrate its utility with a few examples.
The BCB is a joint initiative of the University of Stellenbosch, Copenhagen Business School, and the Wharton School of the University of Pennsylvania. Funding and other support have come from EY, the inaugural sponsor of the Political Risk & Identity Lab, the Norwegian Research Council, the Bay & Paul Foundations, and Amazon Web Services.
The BCB builds from the understanding that business investment and operations are often intertwined with local and national conflict dynamics, and therefore with peacebuilding, development, and state-building goals.
Yet, currently, there is a deep fragmentation of data on the private sector, conflict, and development. Conventional information is often surface-level, slow, and laborious to produce and analyze, difficult to use in cross-context comparisons, or burdensome in access to policymakers, institutions, advocates, and private sector actors. This complicates decision-making, particularly for those who track or manage investment portfolios under political risk in a global setting, and for policymakers who require cross-contextual analyses.
The BCB allows for the ingestion and analysis of a broad range of data relevant to conflict risk and its mitigation. This includes quantitative data, but also textual corpora that can be analyzed through natural language processing and other machine and artificial learning techniques. Data is organized both geographically and temporally, allowing for highly granular snapshots of conditions on the ground.
We call the BCB a “barometer” because it allows us to assess conflict pressures for a particular place and time. Within a corpus of qualitative data compiled by the research team from third-party sources—including media but also corporate, think tank, and advocacy reports—the BCB tracks the incidence of words or phrases associated with conflict events, conflict risk, and moderators associated with conflict intensity. These lexigraphic measures are supplemented with third-party analysis of media reports within 50km provided by the Global Database on Environment, Language, and Tone (GDELT).
The BCB measures conflict on a scale that ranges from violent attacks on the one side to defense against violence placing an actor at risk of harm with most observations in the middle range from criticisms and demands to praise and financial contributions. Conflict and cooperation events across this scale can be associated with particular actors (e.g., businesses, unions, or government), themes (e.g., land use or ethnic favoritism), or combinations thereof (i.e., conflict between an NGO and a firm over environmental impacts).
We can perform this analysis for a particular set of geocoordinates (e.g., for a highly granular perspective on a particular investment), within or across countries (e.g., for a time-series snapshot of the private sector and conflict in general), within or across sectors (e.g., to assess whether particular types of investments are more or less conflict-prone), or for a particular portfolio of interest (e.g., all European Bank for Reconstruction and Development projects; all mines operated by a particular company).
Furthermore, granular time series allow us to assess whether conflict pressures are rising or falling for a particular project, place, region, or sector over hours, days, months, or years.
The BCB thus enables a far more granular assessment of social and political risk than financial actors and others can typically consider using country and industry metrics.
In the future, we will provide a public portal for less technical users. This will allow for more intuitive access to freely available data as well as analyses that draw from a pre-populated library of logical queries and simplified data presentation and visualization.
We are currently developing a research portal that will allow for ingestion of user data for integration within the BCB’s relational database, more complex custom queries and analyses, greater flexibility of inputs and outputs for a wider range of datasets, and management of user licenses for access to proprietary data sources.
Taken together, the BCB dramatically lowers the barriers for interested parties to explore and gain an understanding of the dynamics of the private sector, conflict, and peaceful development—in terms of technological sophistication, technology management, time, and human and financial resources.
The BCB is based on a ground-breaking text-processing engine that facilitates the identification of the location, sector, and timeframe to which a particular document refers while performing textual queries (and suggesting improvements for users) that allow conflict sentiment to be assessed for documents in each corpus against standard conflict measures or custom queries. With the further development of the BCB, researchers and practitioners will be able to use this tool to understand not only the levels of conflict associated with countries and regions but the dynamics of risk over time, allowing for the accurate analyses of future trends and the understanding of the development of political risk through large-scale data analysis.
Applications of the Business and Conflict Barometer: A Political Risk Management Dashboard
The BCB can be used to create a dashboard that identifies flashpoints of political risk for specific times, countries, and sectors like Health Care, which is the most conflict-prone sector globally as a result of the COVID-19 pandemic.
Through this dashboard, we see that regions of large-scale political conflict, such as the Middle East, perform significantly worse than any other region. They consistently demonstrate higher conflict pressure with the textual analysis tool, often picking up keywords such as “fragility,” “armed conflict,” and “peace operations.”
The BCB was used to evaluate country-level conflict in Lebanon in September 2021, during which period it demonstrated a mix of neutral and positive scores in most sectors. However, the “Financials” sector saw a conflict-cooperation score of -3.3 indicative of Lebanon’s battle with one of the worst economic crises on record, resulting in Lebanon’s currency crashing by 90% since 2019 and severe import shortages. In September 2021, despite a lack of fuel imports due to the currency crash, the new Lebanese government further raised gasoline prices. Although initially promised by the government, the central bank stated that it could no longer afford to provide dollars for fuel at the subsidized rate. The increase in risk for the investment and trade climate in the country during the month has led the IMF to recommend that “Lebanon unify the multiple exchange rates along with other steps including the central bank audit.”
Principal Investigators
Brian Ganson, Professor and Head of the Centre on Conflict & Collaboration at Stellenbosch Business School, and Professor, Stellenbosch University School for Data Science and Computational Thinking; Fellow, Wharton Political Risk and Identity Lab
Witold Henisz, Deloitte & Touche Professor of Management; Vice-Dean and Faculty Director, ESG Initiative, the Wharton School, University of Pennsylvania
Anne Jamison, Assistant Professor, Department of International Economics, Government & Business at Copenhagen Business School; Fellow, Wharton Political Risk and Identity Lab