Afghans Don’t Need U.S. Troops. They Need Islands of Stability.

Here’s how the Biden administration can prevent chaos in Afghanistan, even after it withdraws.

By Michael F. Harsch, a visiting assistant professor of international relations at Boston University and a visiting scholar with Harvard University’s Weatherhead Scholars Program, and Taylor Whitsell, a student at Harvard College and a research assistant titled “Islands of Stability in Fragile States,” led by Michael F. Harsch.
An Afghan man feeds pigeons.
An Afghan man belonging to the Uzbek ethnic group feeds pigeons in Mazar-i-Sharif, the capital of the northern Afghan Balkh province, on June 2, 2007. SHAH MARAI/AFP via Getty Images

Last week, U.S. President Joe Biden announced the remaining 3,500 U.S. forces in Afghanistan will be withdrawn by Sept. 11—a decision that sets a clear date for the United States’ military exit. What remains unclear is the administration’s plan for ensuring sufficient stability in Afghanistan without a U.S. troop presence. As September approaches, Biden may face new questions about how the last U.S. forces can leave Afghanistan without leaving chaos behind. Fortunately, there are new answers he can offer.

The U.S. policy in fragile, conflict-affected states is often perceived as a binary choice between either total retrenchment or the largely discredited top-down state-building of recent decades. A multiyear research project on islands of stability suggests a middle ground: a region-driven approach to stabilizing fragile states that could allow the United States to help expand local peace and prosperity in Afghanistan, even amid ongoing conflict, without requiring it to have boots on the ground.

Islands of stability are regions with relatively high levels of security and public service provisions in otherwise fragile, conflict-affected states. Existing islands in states like Afghanistan, Iraq, Syria, and Somalia have demonstrated remarkable resilience in the face of adversity and can provide an important source of stability for the central state. They also accomplish what outside interventions typically fail to promote at the central-state level: a monopoly on violence through long-term cooperation between the government and its citizens.

A region-driven strategy to stabilization would promote the creation and consolidation of such islands of stability. A core element of this strategy would be supporting decentralization efforts and advocating for the direct election of governors in Afghanistan. Islands of stability are more likely to flourish when local citizens have a say and can hold leaders accountable and when provincial governments possess the necessary authority, resources, and popular support to maintain security and deliver basic services. In addition, targeted investments in infrastructure and education in these regions could create positive spillover effects on neighboring regions.

When and why do islands of stability emerge? During interviews in Afghanistan, Iraq, and Somalia, we found local political elites matter greatly. Bounded rulers—who belong to a local majority group that is a national minority and, therefore, possess limited political upward mobility—govern with longer time horizons and have an interest in striking a social compact with their population: long-term protection and public services in return for cooperation with the regional government. Hence, bounded rulers tend to improve security and basic services while in other similar regions, the absence of such leaders is associated with worse outcomes. These leaders also tend to rely on more inclusive social compacts, which extend greater protection and opportunities to women and ethnic minorities than in other regions.

In Afghanistan, which has been ruled by Pashtuns since the 18th century, an example of a bounded leader is Habiba Sarabi, an ethnic Hazara who served as governor of Bamyan from 2005 to 2013 and was the first Afghan woman to head any province.

Another prominent example is ethnic Tajik Atta Mohammad Noor, who served as governor of Afghanistan’s Balkh Province from 2004 to 2018. Realizing his low odds of winning the Afghan presidency, Noor prioritized investing in security, infrastructure, and education to maintain the local population’s long-term support. Offering protection and services enabled Noor to build popular trust and hold the Taliban at bay. Balkh Province’s relative success contrasts with neighboring Kunduz Province, which has been dominated by non-bounded Pashtun appointees and was overrun twice by the Taliban.

Our findings suggest that creating security at the regional level is a feasible and sustainable approach. Unlike peaceful villages that remain highly vulnerable to outside attacks, provinces like Balkh are large enough to defend themselves against most domestic threats; at the same time, they are small enough to enable local accountability and political representation. In the presence of long-standing, bounded leaders, the local population will be ready to share information with the provincial government. As a senior member of the provincial council, whose name we decided to keep confidential given the current political uncertainty in Afghanistan, told us, “except for a fringe minority, the majority of the people … cooperate with Balkh’s government. … When they see anything suspicious, they report it to the relevant authorities immediately.” This allows the authorities to counter threats effectively. An influential local academic, whose identity we are also keeping confidential, pointed out that as a result of this steady information flow, security forces are “able to identify the location … of insurgent groups who are active in Balkh. Therefore, whenever even a small destructive activity takes place, it is clear where the source of the activity lies.”

How can the United States apply the lessons learned from islands of stability in the context of withdrawal from Afghanistan? Going forward, the United States could focus on three areas: decentralization, long-term aid, and disincentivizing foreign subversion.

First, in the context of peace negotiations between the Afghan government and the Taliban, the United States should push for meaningful decentralization. In the current system, governors are appointed by the Afghan president. These governors do not tend to be locally bounded, operating with short time horizons and little local accountability. Turning Afghan governors into elected officials would help ameliorate this problem. Devolution would also reduce the power of the presidency and thereby likely decrease competition over what is currently the paramount political price to capture. This could facilitate a peace agreement.

Although Afghan President Ashraf Ghani will not easily agree to such a change, it is important to keep in mind that even in the absence of forces on the ground, the United States still holds significant sway over the conflict parties. The Afghan government will continue to heavily depend on U.S. financial, diplomatic, and intelligence support, and the United States can now more credibly threaten to walk away, which arguably increases its leverage to push for reform.

Second, donors could create stronger incentives for local leaders to operate with longer time horizons by conducting more long-term aid projects. A promising avenue for reducing conflict is to increase social contact between these islands and neighboring regions by building roads, railways, and stronger communication networks and by improving the quality and accessibility of higher education. Balkh Province, for instance, has several public and private universities, which already attract students from other parts of the country. A local United Nations official, who was not authorized to speak with us on record, suggested that Balkh Province could be transformed into a “major hub for the national economy” and called the province “good raw material for a federal state.”

Third, as islands remain vulnerable to outside interference, a careful and consistent international approach toward these regions and their neighborhoods is required. A U.N.-sponsored conference on Afghanistan could lead to mutual restraint by regional powers. If such an agreement is not possible, the United States should stand firm if external powers, such as Pakistan, continue to embark on military adventurism in Afghanistan. It must make clear that such behavior will have consequences, whether that is international condemnation, the suspension of military aid and exports, or sanctions.

A region-driven approach contrasts with failed attempts by foreign forces to create local “ink spots” of government control in Afghanistan, including the delivery of “government in a box” to newly captured areas. Unlike these, islands of stability already exist or can emerge under local leadership. This stabilization process will be slower yet is more likely to result in lasting change than policies imposed by foreigners on a hostile population.

Balkh Province isn’t Bavaria, and local leaders and conditions in Afghanistan should not be romanticized. Still, islands of stability offer a path toward cultivating reasonably effective, inclusive, and accountable governments within a fragile state. They may represent the best hope for a more effective U.S. approach to stabilizing Afghanistan and other conflict-ridden regions of the world.

Michael F. Harsch is a visiting assistant professor of international relations at Boston University and a visiting scholar with Harvard University’s Weatherhead Scholars Program.

Taylor Whitsell is a student at Harvard College and a research assistant titled “Islands of Stability in Fragile States,” led by Michael F. Harsch.