You can’t win with civil wars

Barbara F. Walter, an authority on civil wars, is an associate professor of political science in the School of International Relations and Pacific Studies at UC San Diego.

Since the beginning of the Iraq war, President Bush has made it very clear that we will stay in that country for as long as it takes to get the job done, and that the United States will prevail in the end. This mantra allows the president to avoid admitting failure, but it ignores everything we’ve learned about civil wars since World War II.

The approximately 125 civil wars -- conflicts involving a government and rebels that produce at least 1,000 battle deaths -- since 1945 tell us several things: The civil war in Iraq will drag on for many more years; it will end in a decisive victory for either the Shiites or the Sunnis, not in a compromise settlement; and the weaker side will never sign a settlement or lay down its arms because it has no way to enforce the terms.

Civil wars don’t end quickly. The average length of all civil wars since 1945 is 10 years. Conflicts in Burma, Angola, India, the Philippines, Chad and Colombia have lasted more than 30 years. Wars in Afghanistan, Cambodia, Lebanon, Sudan and Peru have lasted more than 15 years. Even Iraq’s previous civil war, fought against the Kurds, lasted 14 years.

This suggests that, historically speaking, Iraq’s current civil war could be in its early stages, with nothing to suggest that it will be a short, easy war.

Another lesson from history is that the greater the number of factions involved in a civil war, the longer it is likely to persist. Iraq simply has too many factions, with too much outside support, to come to a compromise settlement now. Not only is there no Shiite or Sunni who can speak for all of his side’s factions, but the parliament seems incapable of stopping the violence between these groups.


This is the problem that Lebanon experienced during its civil war from 1975 to 1991, when multiple competing militias engaged in shifting sectarian violence. It is also the reason why that war was so difficult to end.

Civil wars rarely end in negotiated settlements. In research for a book on the topic, I found that 76% of civil wars between 1945 and 2005 ended only after one side had defeated all others. Only 24% ended in some form of negotiated solution. This suggests that the war in Iraq will not end at the bargaining table but on the battlefield.

True, many Iraqi factions -- including Shiite and Sunni groups -- would prefer negotiation to protracted conflict. After all, political power and oil revenues can be divided. The crux of the problem is that there is no way to enforce any agreement that is reached. How do you convince the Sunnis that they will be able to prevent their enemy from establishing an Iranian-style theocracy once the Sunnis disarm? You can’t. They are a minority of the population and live in the most oil-poor areas, and that makes them vulnerable to exploitation over time.

American officials are talking about creating a power-sharing agreement among the three main ethnic groups because they dare not acknowledge the coming disaster. The push for a constitutional compromise is naive and without any basis in history or realpolitik.

The two most plausible outcomes are a radical Islamic government established by a Shiite faction and supported by Iran, or a Sunni government with ties to Al Qaeda.

One of the things learned over the last 60 years is that peace settlements in civil wars only work when backed by a third party willing to enforce the terms and to protect the weaker side from exploitation. This is why 40,000 NATO troops were stationed in Bosnia after the 1995 Dayton accords were signed, and why the accords in Rwanda eventually fell apart.

The problem in Iraq is that no third party is likely to be willing to guarantee any settlement that is reached. Nobody believes that the United States will stay in Iraq much beyond 2009, or that the Europeans or the United Nations will step in when the United States leaves.

What does this mean for U.S. involvement? One conclusion would be that if we don’t plan to stay for a very long time in Iraq, there is no added benefit in staying a few extra years. At this point, the longer we stay in Iraq, the more American soldiers will be killed and the more likely our presence will help Al Qaeda recruit more supporters.

This is not something that policymakers in Washington want to hear. That’s because the government that inevitably emerges in Iraq is likely to be one that depends on the support of radical elements within the Middle East, not on the support or goodwill of the United States.

Thus, the sad irony of the civil war in Iraq is that by deposing Saddam Hussein, we’ve created a situation that is likely to remind us of why we supported him for so long in the first place.