The Economic Cost of Conflict: An Example of Regression Analysis

Did you every wonder what a war costs, in economic terms, per day? Such calculations have been made retroactively in some cases; for example, the cost of WW2 or the the invasion of Iraq. But how about estimating the cost for future conflicts? Even small conflicts can affect trade and productivity. Larger conflicts may affect education and public health - issues that may have an effect 10 years down the line. Putting a dollar figure on conflicts may help those at the UN or other places involved in Peace Building to demonstrate that wars are perhaps not the best way to resolve a conflict.

To tackle this question, we analyzed publicly available data on conflict and socio-economic indicators for 1960-2010, using datasets from the Center for Systemic Peace (CSP) and the World Bank (WB).

The associations we found were very strong - at major inflection points for conflict, economic indicators invariably followed - sometimes with some dealy but often almost immediately. We performed five detailed cases studies, looking in detail at various indicators - not only trade and GDP but also at agricultural production, school enrollment, gender parity in schools and R&D investments. While many specific observations came out of these cases studies, the general association was clear throughout. So clear in fact that we ended up making a mock calculator to estimate the cost of a low grade (level 2 in CSP terms) conflict per day. 

.

These numbers are of course not realiable enough for the calculator to be used for actual estimates in dollars, But it served to make a point: how computer-assisted regression analysis can provide us with insights that we have never had before.

Contributors to this project: Lambert Hogenhout, Marina Drus

Note: The views expressed herein are those of the author and do not necessarily reflect the views of the United Nations.