While shifting to a light-footprint approach, the Obama administration did not make an equivalent investment in nonmilitary counterterrorism approaches. Through 2015, democracy spending by the Obama administration actually shrunk by significant portions. It was only in the Fiscal Year 2017 budget that the administration reversed course and requested more resources for democracy programming. Obama did make a big push in his second term to elevate “countering violent extremism” (CVE) as a major programmatic area of practice, hoping to shift the focus to the long-term drivers of radicalization and violent extremism. The administration convened an international CVE summit, established an interagency CVE task force and refocused the State Department’s Bureau of Counterterrorism on violent extremism. In 2016, the Obama administration requested $187 million for State Department programs and policies to counter violent extremism—more than twice as much as in 2015.

Yet the CVE agenda brought varying results. Domestically, some civil-rights groups criticized the CVE agenda for unfairly stigmatizing American Muslim communities. On a broader level, CVE programs suffered from a lack of rigorous empirical knowledge about individual and community-level drivers of radicalization in different contexts. In addition, CVE efforts were largely divorced from hard-security approaches, thereby diminishing their influence. As Larry Attree writes, for CVE programs to have a real effect, they cannot “merely go alongside problematic military and rule of law approaches. CVE will only work if it actually stands to change the tactics used by military and criminal justice actors.” This type of integration remained more aspirational than actual.

There was also an uneasy tension between traditional development and governance programs and CVE-specific programming. Particularly for development practitioners, the first-order concern is to respond to locally driven needs and priorities, and to determine how the international community can best support community goals. In contrast, CVE programming asks a fundamentally different question: to what extent can programming in local communities help achieve U.S. counterterrorism objectives? Such an approach risks overlooking local drivers of instability not directly related to extremism, allowing conflicts and grievances to fester. Additionally, CVE programs’ heavy focus on combating extremist ideologies meant that they often gave government actors a pass—even though state security forces are key drivers of insecurity.

AS THE Trump administration proceeds with the complex challenges of counterterrorism, it should bear in mind five lessons.

First, protecting civilians is not just a moral obligation derived from international human-rights and humanitarian law; it is also smart strategy. Violent extremist groups are adept at exploiting local outrage in response to civilian casualties. As dozens of former national-security officials wrote to Secretary of Defense James Mattis in an open letter in March, “Even small numbers of unintentional civilian deaths or injuries . . . can cause significant strategic setbacks.” Civilian casualties alienate local populations, reduce their willingness to cooperate with international forces and foster new grievances that can push individuals toward violence. President Trump should refrain from dismantling Obama’s legal and policy framework governing air strikes and civilian protection. While the Presidential Policy Guidance could be amended to facilitate accelerated prestrike deliberations, these shifts should not come at the expense of interagency input and civilian-protection standards that go beyond the criteria set out in international humanitarian law.

Second, providing greater transparency about U.S. air strikes is critical. The Trump administration should reinforce and bolster President Obama’s commitments to improving clarity around U.S. air strikes. In the face of public backlash over increasing civilian deaths, it may be tempting to once again veil the program in secrecy. This temptation should be resisted. The absence of independent investigations and data plays into the hands of terrorist groups, which can capitalize on civilian populations’ fear and uncertainty and misrepresent casualty numbers. Moreover, discrepancies between official U.S. government reporting on civilian deaths and numbers reported from independent outside sources undermine U.S. credibility, and make it harder to ensure that other countries will comply with civilian protection principles. If the United States underestimates actual levels of civilian casualties, it risks incorrectly prioritizing civilian protection in future operations and obstructing internal learning to mitigate civilian harm.

Third, conducting air strikes without an accompanying political strategy and civilian engagement will not lead to success. Targeted killings can accomplish key tactical objectives. For example, U.S. drone strikes have reduced the number of core Al Qaeda members in Pakistan, and helped thwart the Islamic State’s further territorial expansion in Iraq and Syria. Yet over the course of the past eight years, the limitations of the light-footprint approach have also become evident. According to Long (2014), decapitation strategies that consist of targeting high-level leaders and key middlemen tend to be less effective against highly institutionalized groups that can quickly absorb shocks to their leadership structure. In these cases, targeted attacks may create temporary disruptions and trigger tactical shifts, without diminishing the organization’s overall strength. In addition, a narrow focus on decapitation risks missing, and inadvertently exacerbating, broader conflict dynamics that allow terrorist groups to thrive. For example, while air strikes have dealt serious blows to AQAP in Yemen, the ongoing civil war has allowed the group to benefit from the local war economy and exploit growing sectarianism at the grassroots level.

Fourth, unconditionally embracing repressive governments with poor human-rights records comes with a high cost. Over the long term, these regimes are brittle and prone to collapse. Their regional interests often do not or only partially align with U.S. counterterrorism and stabilization objectives—and even if they do, partner governments may lack the capacity to effectively absorb U.S. assistance. For example, a recent RAND Corporation study found that U.S. military aid aimed at reducing fragility tends to be much less effective in regions with weak state institutions, low state reach and autocratic regimes. Many partner countries advance U.S. counterterrorism objectives in some ways while blocking them in others. The challenge is, thus, to accurately assess partner interests and incentives to maximize gains and avoid doing further harm. In the most extreme cases, U.S. security assistance can prop up governments and security forces that breed further extremism and insecurity. As former NSC Senior Director for Counterterrorism Luke Hartig remarked to us in an interview, “The reality is that we have a very spotty record on military partnerships sustainably mitigating terrorist threats, much less promoting the development, better governance and the rule of law.”

The Trump administration should prioritize strengthening partner accountability. This involves improved intelligence and analysis of the political, economic and social dynamics of partner countries, and how they affect U.S. strategic objectives. Security-sector assistance should take into account the recipients’ domestic and regional political interests, rather than follow an apolitical capacity-building model. The United States should condition training, support and arms transfers on clear benchmarks linked to partner compliance with civilian-protection standards and international humanitarian law. This will likely require withholding or reducing assistance in case of underperformance. If the United States lacks the leverage to press for meaningful security-sector and governance reforms, then it should not be afraid to walk away from the security relationship. In such cases, more U.S. assistance is unlikely to solve underlying tensions, and may instead fuel corruption and insecurity.

Fifth, for U.S. counterterrorism strategy to be effective beyond the near term, the United States needs to support the development of legitimate governance institutions in regions and countries prone to terrorism. Such assistance should go beyond training security forces. Instead, it should include support for meaningful reforms of political institutions to mitigate corruption and bolster inclusive participation. Institution building is a long-term process, and the influence of external actors tends to be limited. Yet by adapting a “security first” approach that disregards governance challenges as secondary objectives, the United States risks further exacerbating the problem.

For example, in Iraq, selective and politicized governance and the systematic torture of Sunni Iraqis by Shia security forces undermined the government and facilitated the Islamic State’s capture of western Iraq. Likewise, gross human-rights abuses propagated by the Nigerian government in response to the Boko Haram insurgency have arguably bolstered the group’s standing and undermined government efforts to retake and stabilize the northeast. In countries like Yemen and Libya, effectively tackling terrorist threats will require renewed efforts to attain political settlements to ongoing conflict. If the Trump administration prevails in its plan to cut governance and development aid and slash the State Department’s budget, the United States will weaken essential tools in the long-term fight against violent extremism. Not surprisingly, many high-ranking military officers have spoken out against the proposed cuts, noting that the military needs “strong civilian partners in the battle against the drivers of extremism—lack of opportunity, insecurity, injustice, and hopelessness.”

The nine months of Trump’s presidency have witnessed some showy moments—Tomahawk missiles raining down on a Syrian airfield, the detonation of the “Mother of All Bombs” in a distant corner of Afghanistan. There has been less evidence of a serious commitment and strategy to truly reduce the threat of terrorism. In fact, signs have pointed in the opposite direction—toward a diminished emphasis on diplomacy and development, and a less rigorous adherence to civilian protection. As Trump’s national-security team continues to take shape, they should pay close attention to the lessons of past presidencies and the accumulated evidence from sixteen years of fighting terrorism.