Working Paper
The Cost of Failing States and the Limits to Sovereignty

In this paper, we estimate the costs of state failure, both for the failing state itself and for its neighbours. In our analysis, the cost of failure arises from two distinct sources: organized violence due to the incapacity of the state to ensure its own citizens' security and low quality of regulation and public goods due to poor governance. To estimate the cost of failure, we proceed in two steps. First we estimate the annual loss of growth induced by state failure. Then we cumulate this loss over time, taking into account the chances that each year a failing state will exit this status. Our growth estimations suggest that a failing state at peace loses 2.6 percentage points of growth per year, while violence induces a further loss of 1.6 percentage points of growth per year. From this, we deduce that states lose around 5 times their initial GDP due to bad governance and a further 0.65 times their initial GDP due to large-scale violence. Using the average GDP of a failing state and taking into account the fact that there are on average 23 failing states, the total cost of failure for the population of these countries, expressed as a net present value, is around US$784 billion (US$39 billion per year). We also find that the cost of failure for the neighbours is similar whether the country is at peace or at war. Neighbours lose around 0.6 percentage points of growth each year. Cumulating over time and taking into account the fact that a typical failing state has around 3.5 neighbours, we find that the total loss for the neighbours is around US$4,730 billion, expressed as a NPV. The annual cost to neighbours is around US$237 billion Our results suggest that the total annual cost of state failure is very large; for example, it far exceeds expenditures on global aid. Moreover, around 80 per cent of the cost is borne by neighbours of failing states: failing states are costly mainly because they inflict externalities on others.