In February 2010, the United States and its allies launched the largest offensive in Afghanistan since the overthrow of the Taliban regime. Its primary goal was to help the struggling Afghan government gain control over Marjah, a remote Taliban stronghold in the southern province of Helmand. It was also a test case for a bold new state-building tactic: delivering security, health services, education, infrastructure, and a handpicked new district governor to the local population as a “government-in-a-box.” The Taliban was driven into the countryside, but the operation soon encountered unexpected problems. The local population mistrusted their dispatched rulers, the inexperienced administration was unable to deliver the public services promised, and reconstruction efforts remained sluggish. The new governor, a stranger to the region who had spent 15 years in Germany, turned out to have a criminal record for stabbing his son. Today, most of the gains made by the U.S. military in Helmand have been lost and the Taliban controls large areas of the province.

The unsuccessful intervention in Marjah is just one example, albeit an extreme one, of liberal state building’s wider failure as the dominant strategy for remaking war-torn countries. Championed by the United States and its allies and typically implemented by the United Nations, the policy aims to transform fragile states into stable democracies by strengthening and expanding the central government’s authority. Yet the approach ignores that central governments frequently lack even the most basic capacity (not to mention legitimacy) to rule their countries. It’s no surprise, then, that most major state-building interventions from Afghanistan to Iraq and Somalia have failed spectacularly.

Dealing more effectively with state collapse will be a key foreign policy priority for U.S. President-elect Donald Trump. Though on the campaign trail he often advocated disengagement from fragile countries, that policy would be shortsighted. In the absence of functioning governments, violent conflict, poverty, refugee crises, and extremist movements may fester and pose serious threats to U.S. security and economic interests.

In Afghanistan, Iraq, and Somalia—three of the world’s most fragile states—a better alternative to liberal state building already exists. Each country has its own “island of stability,” a major region that defies the greater country’s generally negative trajectory: Afghanistan’s northern province of Balkh, Iraqi Kurdistan, and Somalia’s northwest region of Somaliland. In these areas, local rulers provide their populations with surprisingly high levels of security and public services. Some have even established democratic institutions. As strong and relatively autonomous entities, they function as anchors of national stability for states threatened by radical groups such as the Taliban, the Islamic State (also known as ISIS), and al Shabab. Creating conducive conditions for islands of stability would be a more feasible and sustainable approach to aiding fragile countries than either liberal state building or disengagement.

A person on stilts during a parade to mark the 24th self-declared independence day for the breakaway Somaliland nation in the capital Hargeysa, May 2015. 
Feisal Omar / REUTERS


Since the collapse of the Soviet Union, the international community has mainly focused on the challenge, as the political scientist Francis Fukuyama put it, of “getting to Denmark”: transforming unstable, poor countries into peaceful and prosperous democracies. In the 1990s, the dominant approach was rapid liberalization through elections and sweeping economic reforms. Yet the competitive nature of both democracy and a market economy proved ill suited for fragile states and often resulted in a return to armed conflict, such as in Angola and Rwanda.

By the late 1990s, the creation of national institutions—from professional security forces to formal courts and human rights commissions—was increasingly seen as a prerequisite for successful liberalization. This even more ambitious version of liberal state building has since become the dominant approach.

Yet after two and a half decades of unsuccessful engagements, from Haiti to South Sudan, there is a growing consensus in Washington that the United States needs to change its approach to fragile countries. The conventional wisdom is that interventions should be more selective and better prepared. Yet such changes are unlikely to improve liberal state building’s extremely poor track record of success. Of the 25 countries in which the United Nations has mandated major peace operations since 1989, only two—Croatia and Namibia—have emerged as liberal democracies.

Consider also: In the course of what is now the longest war in U.S. history, the United States has spent more money on rebuilding Afghanistan than it spent on the entire Marshall Plan to reconstruct Europe after World War II. Fifteen years after the fall of the Taliban regime, the results of one of the costliest exercises in state building are sobering: the Taliban and other armed groups are on the rise, civilian casualties have reached a record level, the Afghan central government is politically deadlocked, and there are widespread allegations of corruption against government officials.

The United States has spent more money on rebuilding Afghanistan than it spent on the entire Marshall Plan to reconstruct Europe after World War II.

In Afghanistan and beyond, it has become clear that liberal state building is an expensive and failing strategy. Many now conclude that engineering in just a few years what took centuries in western Europe amounts to an impossible dream.


Islands of stability in fragile states demonstrate that local rule can result in relative security and prosperity. According to the nonprofit Fund for Peace, Afghanistan, Iraq, and Somalia all ranked in 2016 among the most fragile states in the world, with Somalia topping the list. Yet Afghanistan’s Balkh Province has experienced a remarkable recovery and remains one of the last pockets of relative peace in the country. Iraq’s autonomous Kurdish region has become the most stable and economically thriving part of the country since the 2003 U.S. invasion. And in the northwest of Somalia, Somaliland has developed into a relatively peaceful and democratically governed entity after its self-declared independence in 1991.

All three regions provide their populations with relatively high levels of security and public services. Data from Uppsala University suggest that between 2002 and 2015, Balkh’s rate of conflict deaths was one-sixth of the national average. In Somaliland during the same time period, the rate was just one-20th of the national average. The case of Kurdistan is similarly striking: the U.S. military recorded hardly any significant security incidents in the region, even at the height of the Iraqi civil war, which in 2006–07 was one of the most violent conflicts in the world. In May 2014, the Economist Intelligence Unit estimated that the region’s reported level of theft was lower than Singapore’s. Since 2014, the Kurdish security forces have withstood and pushed back ISIS with the help of an international coalition.

In addition, these regions boast better public services than almost anywhere else in the country. According to UN and World Bank estimates, in Balkh, from 2007 to 2011, access to drinking water and skilled birth attendants more than doubled; in Iraqi Kurdistan in 2011, 97 percent of the population had access to safe drinking water and 96 percent of children attended primary school. In Somaliland in 2006, primary school enrollment, though low by international standards, was twice Somalia’s national average. And these islands of stability have been able to maintain or further improve these indicators in recent years.


It is tempting but misguided to reduce these regions’ successes to fortunate circumstances. Distance and difficult terrain remove each from its respective capital, providing them with a degree of autonomy; their access to regional and international trade networks generates revenues and facilitates foreign investment. Yet even when Balkh, Kurdistan, and Somaliland are compared with regions within their countries that have similar sizes, geographic positions, and access to trade routes, they still stand out in terms of security and public service delivery. Nor did they get a head start on state building: Balkh was brutally governed by the Taliban in the years prior to the U.S. intervention; Iraqi Kurdistan faced violent oppression and systematic underinvestment during the regime of Saddam Hussein; and in Somaliland during the 1980s, Somali dictator Siad Barre attempted to crush a rebel movement by killing tens of thousands and destroying Hargeisa, the region’s capital.

The comparison of local rule in Afghanistan, Iraq, and Somalia shows that the most successful leaders are “locally bounded” by their ethnic, religious, or clan identity. They rise to power from within the community, and their identity uniquely qualifies them for effective leadership of their community. Crucially, their identity also disqualifies them from leading the country’s central government or other major regions. As a result, these rulers govern with a long time horizon. They tend to prefer taxation to looting their constituencies and, in return, provide the local population with security and public services.

To be sure, local rulers should not be romanticized, but recall the economist Mancur Olson’s famous argument that people living under anarchy are better off if subject to a stationary rather than a roving bandit. Stationary rulers have a long time horizon and, as a result, tend to invest in local stability and institutions. The available evidence confirms Olson’s claim. In Afghanistan, former President Hamid Karzai consolidated his power by appointing five or more different governors to most provinces between 2002 and 2014. By contrast, Balkh’s governor, Atta Muhammad Noor, an ethnic Tajik from Balkh Province and a former military commander of the Northern Alliance against the Taliban, has held the governorship there since 2004. Noor mainly governs through informal networks, coercion, and co-optation, although he has tolerated a comparatively free press and provincial council elections. Under his rule, the province’s infrastructure significantly improved, cross-border trade flourished, and poppy cultivation dropped. Although he faces serious allegations of repression against Pashtun communities, Noor maintains a strong local support base. Yet “the son of Balkh” lacks nationwide support in a country that has been ruled almost exclusively by Pashtun leaders since the eighteenth century. He thus has stronger incentives to propel his constituency toward longer-term stability and prosperity than short-term appointees.

Atta Mohammad Noor, governor of the Balkh province, speaks during an interview in Kabul, Afghanistan, May 2016.
Atta Mohammad Noor, governor of the Balkh province, speaks during an interview in Kabul, Afghanistan, May 2016. 
Mohammad Ismail / REUTERS

In the Kurdistan region of Iraq, a power-sharing agreement between two rival Kurdish parties emerged in the late 1990s and was consolidated after 2003. In 2005, the leader of the Kurdistan Democratic Party, Massoud Barzani, was elected by the Kurdish Parliament as president of Kurdistan, while the head of the Patriotic Union of Kurdistan, Jalal Talabani, became president of Iraq, a largely ceremonial position that he held until 2014. As a Kurdish leader is unlikely to become prime minister of Shiite-majority Iraq, he has to maintain strong popular support by governing effectively and strengthening the region’s autonomy. While most power in Kurdistan remains in the hands of the two parties, the rise of a third party called Gorran (Change) has created an increasingly competitive multiparty democratic system.

Somaliland’s post-1991 political system was established in a series of peace conferences, organized by respected clan elders. Under Somaliland’s second president, Muhammad Ibrahim Egal (1993–2002), a new constitution was adopted that introduced political parties and democratic elections. During this transition, Somaliland’s government primarily relied on local revenues rather than foreign aid. Today, Somaliland has a democratically elected president and a bicameral Parliament—composed of an upper House of Elders and an elected lower House of Representatives. Somaliland’s presidents, who typically belong to the northern Isaaq clan, would be neither able nor willing to claim the presidency of Somalia. But they have done well enough in Somaliland. Since 1993, presidents have governed for nearly eight years on average before handing power over democratically and peacefully to a successor. After a controversial delay of its latest parliamentary elections, Somaliland is scheduled to elect a new Parliament and president in spring 2017.


None of the regions examined here is likely to “get to Denmark” anytime soon. Yet in all three, locally bounded rulers have established reasonably effective and accountable public institutions.

One important lesson for the United States is that less foreign support might be more beneficial for fragile states. If governments primarily rely on foreign aid rather than on local tax revenues, political accountability will decline, corruption levels will rise, and leaders’ incentives for delivering public services will decrease. Although extensive yet futile attempts were made to strengthen the central governments in Afghanistan, Iraq, and Somalia, the international footprint in Balkh, Kurdistan, and Somaliland remained relatively light. As a result, these regions have developed greater self-reliance and use available resources more effectively.

One important lesson for the United States is that less foreign support might be more beneficial for fragile states.

The United States should also encourage decentralization and cautiously promote fragile countries’ islands of stability. It could help these regions by facilitating access to trade networks, advocating local autonomy in selecting leaders, and encouraging the longer-term establishment of more inclusive institutions. This approach is unlikely to work in active proxy wars such as those in Syria and Yemen, where external actors are fueling the armed conflict to such an extent that any local stability remains impossible. Yet for Afghanistan, Iraq, and Somalia, strong regions are a source of stability for the state: Balkh has helped to slow the momentum of the Taliban; Iraqi Kurdistan’s military contributions have been vital in containing ISIS; and Somaliland might become crucial in the fight against al Shabab.

The international record of both Trump and new UN Secretary-General António Guterres will be judged in part by their ability to engage more successfully with fragile states. They should seize the opportunity to revisit the current orthodoxy and learn from the surprising successes of these islands of stability.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • MICHAEL F. HARSCH is Assistant Professor of Practice at New York University Abu Dhabi (NYUAD) and a Principal Investigator of the project Islands of Stability in Fragile Countries at NYUAD.
  • More By Michael F. Harsch