Kigali, the peaceful capital city of Rwanda. Our global results confirm those from the the Institute for Economics and Peace analysis of the Rwanda experience, conservative costs of prevention reap 16 to 1 returns against the costs of war. Photo: Flickr / Dylan Walters.
Benjamin Franklin said that “An ounce of prevention is worth a pound of cure.” As the world spends billions on the world’s cures, peacekeeping, post-conflict reconstruction and humanitarian relief in response to the world’s violence, we must ask whether prevention could save those costs? The problem of prevention is that there are no guarantees; we would very much like to avoid wars, but we have no assurance that wars will happen when they are predicted or that we can avoid them with our prevention tools. Unfortunately, prevention is not free – it requires political will and resources. Prevention is certainly worth it in hindsight, but how do we know when it is worth it for the wars of the future?
The beginning of an answer to this question is in the forthcoming joint United Nations–World Bank (UN–WB) study on prevention to which we contributed – the summary report was launched at the UN General Assembly this week. For this report, one of the authors here wrote a background paper, looking at the historical probabilities of countries moving from one state – ‘at peace’, for example – to another state – ‘at war’. To build these probabilities, he examined all developing countries and considered how they transition between different states: at peace; high risk; armed conflict; at war; and recovering. The ‘high risk’ category was developed from a forecasting framework which was built to spot the start of civil wars a year before they occurred. As one would expect, it is very rare for countries to move immediately from ‘at peace’ to ‘at war’; less than 1 per cent do. It is more common for countries that are at ‘high risk’ to have a conflict, but only about 12 per cent experienced political violence on their soil in the following year.
Prevention simulations
These probabilities serve as the basis for simulations that were run in the UN–WB report (see box 1 in the report). These simulations allow us to model the effects if the world community decides to intervene more decisively in ‘high risk’ situations through diplomacy, aid and other preventive measures. Since the success rate of such a prevention system is unknown, three different scenarios are run. For the pessimistic scenario we assume that prevention only works 25 per cent of the time. This leads to weak results in the probability of an outbreak of conflict. But remember, the assumption is that these happen every year and gains can build up over time. For the neutral scenario, we assume that prevention is successful 50 per cent of the time and for the optimistic scenario, we assume that prevention is successful 75 per cent of the time.
Figure 1 shows the number of preventions required in these scenarios. In the first year the model produces ten ‘high risk’ situations and so ten interventions are required in every scenario. However, this number quickly falls because a share of the ‘high risk’ situations are de-escalated. Under the assumption of 50 per cent efficiency on prevention (it only manages to avert war 50 per cent of the time) it would quickly become necessary to intervene only five times per year after 2020. The lower effectiveness in the pessimistic scenario means that more interventions are necessary but fewer outbreaks are prevented.
Figure 1. Prevention actions per year. Graphic: Authors.
Note that the ‘high risk’ situations are identified through forecasting and therefore prone to error. This is one of the problems highlighted above with prevention. Due to this forecasting error, about 9 preventions in 10 would be in vein because no conflict would break out in the following year even in the status quo scenario. On top of this – to remain realistic – we assumed that prevention does not always work.
Still, if we compare the prevention scenarios with a status quo scenario, definitive change is evident. Figure 2 depicts how many additional countries would enjoy peace (at peace or high risk) in the respective prevention scenarios. We can see a steady increase in the number of countries at peace. In the neutral scenario the world would, on average, have at least four additional countries in a state of peace by 2030. Even in the pessimistic scenario there would be peace in two additional countries.
How can a serious prevention effort, even with the forecasting error and the limited effectiveness, lead to such notable changes in the long run? A key aspect is the dynamic change in risk implied by prevention. The start of a civil war is a path that often leads years of open violence and decades of repeated cycles between low intensity armed conflicts, outright civil war and recovery. Given this high persistence of conflict and high risk of relapse, conflict prevention before entering the conflict cycle engenders re-enforcing, long-term, positive dynamics. As more countries are prevented from slipping into violence, the world becomes a more peaceful place.
Figure 2. Comparison to status quo - how many more countries would enjoy peace. Graphic: Authors.
Before we consider the returns on our investment from prevention, it is useful to remember the human costs of conflict and lives that could be saved from prevention. In Figure 3 we show the cumulative number of prevented deaths from the prevention system compared to the status quo. It shows that, by 2030, close to 150,000 lives could be saved. Even in the most pessimistic scenario, we could save more than 75,000 lives with serious prevention. This alone should be reason enough to engage with more systematic and robust prevention strategies.
The returns to prevention can basically be broken down into two parts: saved costs (the costs of reconstruction through foreign aid, the costs of peacekeeping in the recovery phase, humanitarian response, and so forth); and prevented damages (the economic value of lives lost and the cost of foregone economic growth due to conflict). Due to the huge impact that violence has on economic growth, the lost gross domestic product (GDP) plays a key role in the loss estimate. To calculate total costs, we must also consider the additional cost of prevention activities. The scenarios assume a whole range of intervention costs of $100 million, $500 million to $1 Billion per prevention year.
In summary, the neutral scenario assumes that interventions that cost $500 Million per year only reduce the risk of an escalation by 50% – about six percentage points in the model. Still the results are striking: under the neutral set of assumptions, $2 Billion in average spending on four preventions per year, the eventual annualized average savings from avoided conflict is $33 Billion per year. This $33 Billion is net of the savings on prevention and reflects both GDP gains as well as avoiding fatalities and expenditures on peacekeeping and humanitarian response. These numbers are very similar to those arrived at by the Institute for Economics and Peace, estimating that all of the costs of preventive activity in Rwanda since the genocide have reaped 16:1 benefits in the conflicts that have been avoided. 16 to 1 – what a convenient coincidence to find that an ounce of prevention is truly worth a pound of cure.
Benjamin Franklin said that “An ounce of prevention is worth a pound of cure.” As the world spends billions on the world’s cures, peacekeeping, post-conflict reconstruction and humanitarian relief in response to the world’s violence, we must ask whether prevention could save those costs? The problem of prevention is that there are no guarantees; we would very much like to avoid wars, but we have no assurance that wars will happen when they are predicted or that we can avoid them with our prevention tools. Unfortunately, prevention is not free – it requires political will and resources. Prevention is certainly worth it in hindsight, but how do we know when it is worth it for the wars of the future?
The beginning of an answer to this question is in the forthcoming joint United Nations–World Bank (UN–WB) study on prevention to which we contributed – the summary report was launched at the UN General Assembly this week. For this report, one of the authors here wrote a background paper, looking at the historical probabilities of countries moving from one state – ‘at peace’, for example – to another state – ‘at war’. To build these probabilities, he examined all developing countries and considered how they transition between different states: at peace; high risk; armed conflict; at war; and recovering. The ‘high risk’ category was developed from a forecasting framework which was built to spot the start of civil wars a year before they occurred. As one would expect, it is very rare for countries to move immediately from ‘at peace’ to ‘at war’; less than 1 per cent do. It is more common for countries that are at ‘high risk’ to have a conflict, but only about 12 per cent experienced political violence on their soil in the following year.
Prevention simulations
These probabilities serve as the basis for simulations that were run in the UN–WB report (see box 1 in the report). These simulations allow us to model the effects if the world community decides to intervene more decisively in ‘high risk’ situations through diplomacy, aid and other preventive measures. Since the success rate of such a prevention system is unknown, three different scenarios are run. For the pessimistic scenario we assume that prevention only works 25 per cent of the time. This leads to weak results in the probability of an outbreak of conflict. But remember, the assumption is that these happen every year and gains can build up over time. For the neutral scenario, we assume that prevention is successful 50 per cent of the time and for the optimistic scenario, we assume that prevention is successful 75 per cent of the time.
Figure 1 shows the number of preventions required in these scenarios. In the first year the model produces ten ‘high risk’ situations and so ten interventions are required in every scenario. However, this number quickly falls because a share of the ‘high risk’ situations are de-escalated. Under the assumption of 50 per cent efficiency on prevention (it only manages to avert war 50 per cent of the time) it would quickly become necessary to intervene only five times per year after 2020. The lower effectiveness in the pessimistic scenario means that more interventions are necessary but fewer outbreaks are prevented.
Note that the ‘high risk’ situations are identified through forecasting and therefore prone to error. This is one of the problems highlighted above with prevention. Due to this forecasting error, about 9 preventions in 10 would be in vein because no conflict would break out in the following year even in the status quo scenario. On top of this – to remain realistic – we assumed that prevention does not always work.
Still, if we compare the prevention scenarios with a status quo scenario, definitive change is evident. Figure 2 depicts how many additional countries would enjoy peace (at peace or high risk) in the respective prevention scenarios. We can see a steady increase in the number of countries at peace. In the neutral scenario the world would, on average, have at least four additional countries in a state of peace by 2030. Even in the pessimistic scenario there would be peace in two additional countries.
How can a serious prevention effort, even with the forecasting error and the limited effectiveness, lead to such notable changes in the long run? A key aspect is the dynamic change in risk implied by prevention. The start of a civil war is a path that often leads years of open violence and decades of repeated cycles between low intensity armed conflicts, outright civil war and recovery. Given this high persistence of conflict and high risk of relapse, conflict prevention before entering the conflict cycle engenders re-enforcing, long-term, positive dynamics. As more countries are prevented from slipping into violence, the world becomes a more peaceful place.
Before we consider the returns on our investment from prevention, it is useful to remember the human costs of conflict and lives that could be saved from prevention. In Figure 3 we show the cumulative number of prevented deaths from the prevention system compared to the status quo. It shows that, by 2030, close to 150,000 lives could be saved. Even in the most pessimistic scenario, we could save more than 75,000 lives with serious prevention. This alone should be reason enough to engage with more systematic and robust prevention strategies.
How much is prevention worth?
The returns to prevention can basically be broken down into two parts: saved costs (the costs of reconstruction through foreign aid, the costs of peacekeeping in the recovery phase, humanitarian response, and so forth); and prevented damages (the economic value of lives lost and the cost of foregone economic growth due to conflict). Due to the huge impact that violence has on economic growth, the lost gross domestic product (GDP) plays a key role in the loss estimate. To calculate total costs, we must also consider the additional cost of prevention activities. The scenarios assume a whole range of intervention costs of $100 million, $500 million to $1 Billion per prevention year.
In summary, the neutral scenario assumes that interventions that cost $500 Million per year only reduce the risk of an escalation by 50% – about six percentage points in the model. Still the results are striking: under the neutral set of assumptions, $2 Billion in average spending on four preventions per year, the eventual annualized average savings from avoided conflict is $33 Billion per year. This $33 Billion is net of the savings on prevention and reflects both GDP gains as well as avoiding fatalities and expenditures on peacekeeping and humanitarian response. These numbers are very similar to those arrived at by the Institute for Economics and Peace, estimating that all of the costs of preventive activity in Rwanda since the genocide have reaped 16:1 benefits in the conflicts that have been avoided. 16 to 1 – what a convenient coincidence to find that an ounce of prevention is truly worth a pound of cure.
ABOUT THE AUTHOR(S)