Firms Response to Violent Conflicts

By | May 14, 2021

A large fraction of the world’s population live in regions affected by violent conflicts and instability. In 2016, more countries experienced violent conflicts than at any point in the last 30 years (United Nations, 2020). Violent conflicts such as war and terrorism can severely disrupt economic activity and determine the way economic agents behave. The effort to understand the effects of violent conflict at a micro-level has historically focused on individuals, households, and local communities. In particular, researchers have studied the impact of violent conflicts on health, education, and individuals’ attitudes towards risk. Very rarely have researchers focused on the impact on the business sector. 

In our paper ‘Firm Responses to Violent Conflicts’, we study the impact of violent conflicts on firms’ decisions in the context of developing countries using Mozambique as a setting. Episodes of violence are one of the biggest challenges of central and local governments and are a permanent source of instability in the country. Moreover, managers in Mozambique have identified instability as one of the biggest barriers to business activity (Financial Times, 2015). Recently, Total, a multinational oil and gas company, declared force majeure on a multibillion dollar project in Mozambique due to an escalation in attacks by Islamist insurgents (Financial Times, 2021). These challenges have also been documented by global institutions: the World Bank Enterprise Survey (2018) reports “Crime, Thefts and Disorder” as one of the greatest obstacles for firms in Mozambique.  

Such a reality is not circumscribed to Mozambique as other African countries struggle to contain the eruption of violent political conflicts as well. Even in developed economies, including the United States, there have been recent episodes of political violence such as the ones associated with the Black Lives Matter Movement or the 2021 post-election Capitol riot. “American polarization, fear, and rage have grown so great that a recent poll showed that Americans believe we’re two-thirds of the way to a civil war,” reported the Times in October 2020. 

Map

Description automatically generated with medium confidence

We use monthly variation in violent political conflicts at the province level in Mozambique to estimate dynamic treatment effects on firm decisions to purchase inventory. The analysis of inventory purchases is relevant to understand the impact of conflict on business activity and investment, as inventory can represent a significant part of a firm’s working capital investment. Our data permits the study of how firms adjust their inventory purchases at a more granular level and at higher frequency than existing studies. Under the assumption that the effect of conflict on business activity happens mostly in the short-term, these effects are more difficult to estimate using annual financial data, which is typically more readily available for firms. 

Firms might respond to violent conflicts by purchasing less inventory in anticipation of a decrease in consumer demand. This is done for precautionary reasons, as violence might lead to property destruction including storage facilities, or because firms’ operations are disrupted. Inversely, in the short-run, firms might decide to increase their inventories despite an expected decrease in consumer demand if they anticipate the conflict to escalate and disrupt supply chains in the future. Overall, the direction, magnitude, and persistence of these effects are unknown. 

The heterogeneity of violent conflicts’ impact across firms of different sizes in the context of developing countries is also unclear. On the one hand, larger firms are expected to be more sophisticated when responding to demand changes and, for this reason, to manage inventories more actively. On the other hand, smaller firms might be more exposed to violence and crime, and face more severe disruption. In addition, firm size can be correlated with other factors, such as financial constraints, risk management practices, and mark-ups, which may exacerbate the impact of conflict. 

We perform firm-fixed effects regression analysis using monthly purchases of 431 client firms of a large multinational beverage firm on the occurrence of political conflicts at the province-month level. Conflict measures use data from The Armed Conflict Location and Event Data. In this analysis, we control for average temperature, monthly inflation, and GDP quarterly growth, as these variables may be correlated with consumer demand for beverages as well as the occurrence of conflicts.  

We find that firms exposed to conflicts on a given month significantly reduce their inventory purchases in the following month when compared to control firms. Our results are economically relevant: there is a reduction of 11% to 15% in inventory purchases following the occurrence of at least one violent conflict. These results are robust to using different econometric specifications to capture time trends and the seasonality of the consumption of beverages.  

This effect is driven by smaller firms, which experience a 18% to 20% stronger decline when compared to large firms. This finding does not seem consistent with violent conflicts impacting firms through disruptions in supply (e.g., delayed or cancelled deliveries). These disruptions would likely affect both larger and smaller firms because of the common supplier. Similarly, our results suggest that reductions in purchases are not driven by an aggregate decrease in demand at the province level, as only small firms are affected.  

Table

Description automatically generated

Next, we investigate whether the impact persists over time. We find that the negative effect of the occurrence of political conflicts on small firms persists for two months, a period beyond which we do not observe any difference between the affected and unaffected firms. This evidence suggests that the impact of violent conflicts on small firms’ outcomes is significant but short-lived, most likely due to increased uncertainty rather than capital or property destruction. This interpretation is consistent with Bloom (2009), who finds that uncertainty reduces investments because firms assume a “wait and see” attitude. It is also consistent with Abadie and Gardeazabal (2008) who show that terrorism may impact the mobility of productive capital even if it destroys only a small fraction of the stock of capital in a country. We also do not find evidence that firms adjust inventory purchases in anticipation of future conflicts. 

Chart, box and whisker chart

Description automatically generated

We complement the previous analysis with data from The Survey of Mozambican Manufacturing Firms (UEM, KU and UNU-WIDER, 2017) to test the validity of the findings in other sectors of activity. This survey was conducted in 2017 for more than 500 manufacturing firms in 7 Mozambican provinces. By analyzing the manufacturing sector, we extend the analysis from a setting based on finished goods inventory (beverages) to one where intermediate inputs inventory have more relevance. First, we document a negative impact of violent political conflicts on small firms’ total revenue. An increase in conflicts from 2015 to 2016 is associated with a drop in total revenues of 11.7 percentage points. This outcome is important as it measures the economic impact of conflict on firms’ turnover (including potential changes in consumer demand or prices). We also find that an increase in the number of violent political conflicts negatively affects the purchase of intermediate inputs in small firms (21.5 percentage points lower relative to larger firms). Thus, we find consistent evidence that smaller firms’ inventory purchases are more affected by conflicts in both data. The reduction in purchase of intermediate inputs persists after controlling for changes in total revenue to account for the correlation between total input value and product demand. This evidence suggests that firms adjust purchases for precautionary motives and not only due to decreased demand. Last, we find that managers of firms exposed to an increase in violent conflicts are more likely to express intentions to expand to less violent areas. This suggests that managers incorporate the occurrence of conflicts in their strategic decisions, and not only in the most operational ones.  

This study contributes to our knowledge about the economic consequences of violent political conflicts with new evidence on firms’ responses. Understanding the impact of conflict on firm outcomes is a first order topic because the private sector is an important driver of growth in developing countries (OECD, 2007). Moreover, our findings are particularly relevant given the lack of evidence on firms’ investment decisions under uncertainty, particularly in developing countries affected by violent conflicts. 

Claudia Custodio is an Associate Professor of Finance at Imperial College, a Research Affiliate at CEPR, and a Research Associate at the Financial Markets Group and ECGI 

Bernando Mendes is a Research Assistant at Imperial College Business School and a Teaching Assistant at Nova School of Business and Economics 

Diogo Mendes is an Assistant Professor of Finance at Stockholm School of Economics and an affiliated researcher at MISUM 

This post is adapted from their paper “Firm Responses to Violent Conflicts” available on SSRN.  

Leave a Reply

Your email address will not be published. Required fields are marked *