Paper - Belfer Center for Science and International Affairs, Harvard Kennedy School

Partnering to Protect: Reforming US Security Assistance to Reduce Civilian Harm

Download the Full Publication

Executive Summary

Security assistance has long been an important component of American statecraft. The foreign policies of successive administrations have focused on empowering US partners to confront their own security challenges, rather than solving them through American force alone. Today, US foreign and defense policy indicates an intent to work “by, with, and through” partner forces to achieve shared goals. However, the outsourcing of American security objectives comes with a host of potential risks to civilians living in conflict zones or fragile states. Fighting with or relying on local partners—whose interests, priorities, and capabilities may not necessarily align with those of the United States—can complicate or even degrade America’s ability to minimize civilian harm during military operations.

This Policy Analysis Exercise (PAE) examines the causes and consequences of civilian harm in security partnerships, and its implications for US foreign policy. The United States understands the strategic importance of protecting civilians during military operations and has learned valuable lessons from nearly two decades of counterinsurgency efforts in Iraq and Afghanistan. But, American officials have not paid the same careful consideration to the increased risk to civilians when US forces operate “by, with, and through” local partners. When partner forces abuse or kill civilians with “Made in the USA” weapons, training, and support, it damages US interests. US engagement with those forces alienates local populations, reduces US credibility, and ultimately undermines America’s interests in international peace and security.

This report provides analysis and recommendations for a congressional office to lead on the important and timely policy problem of civilian harm in security partnerships. It was informed by extensive open source analysis and interviews with practitioners in the security assistance and civilian protection communities. The report proceeds in five parts. The first section provides background and defines the terms that will be used in the report. Section II explores in greater detail the relationship between civilian harm and US security assistance. Section III performs a gap analysis of congressional tools available to mitigate civilian harm in partnerships. Section IV explores three case studies – assistance to Nigeria to counter Boko Haram, support for the Saudi-led coalition in Yemen, and cooperation with the Philippine armed forces against domestic terrorist groups – to illustrate US tools, policies, and processes to reduce civilian harm in partnerships. The report closes with recommendations for Congress and the executive branch.


Key Findings

Unpacking the complex relationship between security assistance and civilian harm is an essential contribution of this study. When the United States does not fully control, monitor, or understand the actions of actors with whom it is associated on the ground, civilians may pay the price. This report finds that the level of civilian harm perpetrated in partnerships depends on several factors:

  • The partner’s independent capability, capacity, and will to prevent civilian harm
  • The nature of the partner’s control of its security forces
  • The alignment of a partner’s threat perceptions with those of the United States
  • The United States’ relative dependence on the partner for wartime information
  • The United States’ ability and willingness to change partner behavior

Responsibility for this complex policy area falls between the executive and legislative branches. The executive branch develops and implements security assistance policy. But current officials do not have a clear strategy to address the increased risk of civilian harm from partnerships. The executive branch’s traditional focus on preserving bilateral relationships, selling equipment without understanding partners’ capabilities and intentions, and an overreliance on human rights training as a panacea can all contribute to risks to civilians in conflict.



This report provides recommendations for Congress and the executive branch to improve civilian protection in partnerships. These recommendations were developed and selected according to their performance on four key criteria: political feasibility, cost and budgetary considerations, implementation factors, and potential for impact.


To the Congress:

  1. Mandate the Department of Defense (DoD) and Department of State (DoS) to develop a strategy to address the increased risk of civilian harm when US forces work “by, with, and, through” partner forces. The absence of such a strategy sidelines and compartmentalizes considerations of civilian protection in partnerships. The strategy should include a framework to assess, monitor, and evaluate a partner’s commitment to civilian protection and establish the triggers and indicators that would require a re-evaluation of the partnership.
  2. Conduct more consistent, effective, and transparent oversight of the executive branch’s strategy for and implementation of US security assistance policy. Congress should improve notification and reporting requirements to better track which countries are getting what kinds of assistance in order to hold partners accountable for abusive behavior. Congress should also demand data and transparency from the executive branch on how security assistance initiatives advance US objectives and how assistance may contribute to civilian harm. Individual members can play a larger role in this process and bring in new voices and constituencies to inform oversight.
  3. Amend the Leahy laws to close loopholes in interpretation and implementation. Leahy vetting is one of the most important tools to ensure US assistance is not used to perpetrate abuses. However, in their current form, the laws do not sufficiently mitigate partner-caused civilian harm. Congress should amend the Leahy provisions to extend their applicability to Foreign Military Sales and set clearer guidelines about which partner country entities are eligible for assistance.
  4. Exercise greater influence and control over US arms sales strategy, policy, and execution. Congress should clarify the conditions under which it will approve future arms sales and reject or hold sales that are inappropriate for the partner’s needs, capabilities, and intentions. Congress should also extend the notification deadline for congressional disapproval from 30 to 60 days, expand end-use monitoring programs to include an evaluation of how US weapons are used in conflict, and ensure that packages include tailored training and technical assistance on civilian protection.
  5. Increase oversight of training programs on civilian casualties and the law of armed conflict. Training is often viewed as a silver bullet solution to a partner’s civilian protection problems. Congress should require DoD to assess, monitor, and evaluate the impact of such training programs to understand their impact.


To the Executive Branch:

  1. Assess partner capability, capacity, and will to protect civilians and establish civilian harm-related triggers. The establishment of such a framework would enable the United State to enter partnerships with a clear understanding of the partner’s weaknesses and gaps on civilian protection and appropriately calibrate and sequence the level and form of assistance. It would also enable policymakers to adapt or modify existing partnerships when conflict conditions change on the ground.
  2. Place conditions on training, equipment, and other support based on partner forces’ commitment to and performance on civilian protection indicators. The exact impact of conditions on security assistance is uncertain and warrants further research. However, the Departments of State and Defense should limit the provision of unconditional military aid to partners while also identifying opportunities to implement more positive conditionality with partners to incentivize better behavior.
  3. Improve monitoring of civilian harm in partnerships. Documenting and analyzing civilian harm is critical to improving protection. The United States should develop metrics and information channels to independently assess and report on civilian harm caused by partner forces. The United States should also help partners enhance their own capacity to track these metrics.
  4. Improve training for partner forces to enable them to better plan and conduct military operations with civilians in mind. DoD’s human rights training is too academic, overly legalistic, and doctrinal. Training should instead focus on working with partner militaries to devise practical civilian protection plans and policies that will prepare them to respond during a real conflict.
  5. Harmonize messaging on civilian protection across US government stakeholders to partners. The message the United States delivers to partners can be segregated and uncoordinated, with officials in the military, in Washington, and at embassies communicating different priorities. Policymakers should work to convey complementary messages to foreign counterparts on the importance of civilian protection.


Looking Ahead

This report demonstrates how security partnerships in armed conflict are uniquely risky to civilians. Improving policy in this area will require engaged, sustained, and effective congressional leadership. Given the increased role partnerships are playing in US foreign policy, Congress must act now to reform US security assistance policy to reduce and mitigate risks to civilians.

For more information on this publication: Belfer Communications Office
For Academic Citation: Lee, Martha, Alexandra Schmitt and Gabrielle Tarini. “Partnering to Protect: Reforming US Security Assistance to Reduce Civilian Harm.” Paper, Belfer Center for Science and International Affairs, Harvard Kennedy School, June 2019.