Civilians and “By, With, and Through”

Key Issues and Questions Related to Civilian Harm and Security Partnership

CSIS Brief


The Issue

  • Working by, with, and through partners in military operations has become a preferred approach in U.S. security policy. Doing so without uniform controls governing conduct and the use of force can result in real consequences for civilians and compromise mission effectiveness.
  • The real and perceived benefits of partnered operations can include limiting the extent of U.S. involvement and minimizing risk to U.S. personnel, tapping into the unique capacities of national and local forces, and burden sharing of costs, personnel, and assets.
  • The risks of partnered operations can arise from the diffusion of responsibilities and diversion of shared interests and objectives. They may result in civilian casualties, damage to civilian infrastructure, human rights abuses, erosion of U.S. or partner legitimacy, reduction of U.S. domestic support for operations, and long-term humanitarian, economic, governance, and security consequences for civilians.
  • Developing policy and doctrinal guidelines, robust security sector assistance, and transparency and accountability measures for partnered operations are among the steps U.S. policymakers and practitioners can take to mitigate the potential for civilian harm by security partners, in addition to identifying opportunities to limit or suspend partnerships with actors demonstrating poor human rights records or failure to meet international humanitarian law obligations.

Introduction: Context

The U.S. National Security and Defense Strategies, as well as trends in U.S. military operations, reflect a growing preference for working by, with, and through allies and partners to achieve U.S. security objectives. Partnered operations may be pursued to limit deployment of U.S. forces, and therefore minimizing risk to them, and may have the real and perceived benefits of burden sharing of costs, personnel, and assets. While “fighting together” can augment the capacity of any one state acting alone, it can simultaneously complicate—or even degrade—transparency, accountability, and consistency in minimizing and accounting for civilian harm.2 At the same time, partnered operations can better mitigate civilian harm if designed to capitalize on comparative advantages, such as access to local populations, language, and oversight and accountability institutions that regulate security force conduct. Examining the risks to civilians associated with the most common forms of security partnerships, and the underlying dynamics contributing to these risks, is essential to finding the most effective means of addressing them.

Recent conflicts have resulted in devastating impact on civilian populations, especially when hostilities take place in urban areas.3 The cities of Mosul, Iraq, and Raqqa, Syria, will require decades to rebuild critical infrastructure and clear explosive remnants of war. With global displacement at its peak since World War II, parties to conflict must also consider the impact of military operations on civilian populations forced to flee their homes. Civilians in these conflicts experience traumatic injuries, loss of family members, the destruction of homes, water and sanitation systems, power plants, and markets, and the disruption of education, livelihoods, and essential services. Partnered operations may lack clarity on the culpability for civilian casualties and damage to civilian infrastructure, which can erode the legitimacy of the partnership in the eyes of civilians and others.

To mitigate the risks while optimizing the benefits of working by, with, and through security partners, the United States and its partners should evaluate the full range of legal, policy, and operational variables specific to different models of working with partners in military operations that aim to manage risk, leverage partner strengths, and improve performance.4

Afghan and U.S. soldiers talking with a civilian — Ted Aljibe/AFP/Getty Images


Recurring Policy and Operational Themes


Purpose of the Partnership


Outcomes and objectives for each security partnership should be made clear up front in direct dialogue with the partner and the public in the partner country and reviewed over time. When preferences about conduct and accountability to civilians differ between the United States and its partners, the conduct or decisions of either party can dramatically affect the other. The United States should thus evaluate the alignment of interests and expectations with its partners so that it can then identify and mitigate risks with adequate controls. The formation of new partnerships should be informed by how the parameters and expectations of such operations intend to meet the security needs of the affected population.

Policymakers do not always consider in advance what types of partner conduct are unacceptable, nor do they anticipate how the United States may need to calibrate or even withdraw its support for partner operations if certain lines are crossed. Clear and consistent communication with the partner is vital to establish consistent expectations, prevent such lines from being crossed, and mitigate the costs of the United States backing out of the relationship if partner conduct does not improve.


The lack of a common methodology for classified and open source intelligence collection, interpretation, evaluation, dissemination, and adjudication may pose challenges for the United States and its partners in their efforts to mitigate civilian harm. Relying on partner-sourced intelligence for kinetic operations may expose the United States to errors in fact or judgement that lead to civilian harm. Partners may also have incentives or political motivations that differ from the United States and lead to the misidentification of combatants. Questions may also arise from the diffusion of decisionmaking regarding responsibility and accountability when faulty intelligence creates operational or even legal hazard (e.g., Somali drone victims suing the Dutch government for information sharing).

The United States and its partners may also differ in their approach to gathering and assessing third party information (all-source intelligence), or simply fail to establish means of transmitting third party information, which could alter estimates of civilian harm before an operation and complicate efforts to track and assess instances of harm afterward. Finally, in determining when intelligence becomes “actionable” to prompt an operational response, the United States and its partners could differ on the response threshold.

Photo: Matej Divizna/Getty Images


Combined Planning, Operations, and Command and Control

Reliance on local security forces can affect the U.S. cost-benefit analysis for an operation; however, it is unclear to the public and the humanitarian community to what degree civilian harm mitigation factors are prioritized in this analysis. Working by, with, and through local partners may, in fact, increase the risks of escalation and exposure for U.S. forces vis-à-vis adversaries and in U.S. forces’ relationships with the local civilian population. This may be exacerbated by different command and control approaches with partners. By their nature, partnered operations require relinquishing some decisionmaking authority at the tactical and operational levels, diluting the level of control over partner conduct against civilians. This distribution of responsibility for decisionmaking can thus distort incentives for attribution and accountability among partners.

The United States may already employ significant measures to control for risks created in a shared command and control structure, to include ensuring thorough oversight and accountability for the conduct of its partners, and what skills and capacities are most important for U.S. partners to achieve intended military objectives while mitigating the effects of hostilities on civilian populations. However, current U.S. practice in this regard is not well understood outside of U.S. government channels.

Security Sector Assistance for Partners

U.S. approaches to capacity building, “train and equip” programs, and joint exercises merit review as they relate to mitigating civilian harm in complex operating environments. Although training has limited impact on conduct in the absence of accountability structures (including policy, doctrine, and command culture), institution building and advising may take years to accrue meaningful results. Yet, scenario-based exercises, applied training, continuous advising, and institution building are among the best ways to mitigate potential tendencies for or instances of civilian harm “upstream” in the security partnership before they occur.5 When directed to train partner forces in the absence of structural or institutional controls on conduct, U.S. forces may struggle to balance the gains realistically possible in the short term against the heightened risk to military personnel and civilians in the meantime. Finally, the pedagogical aspect of civilian harm mitigation seems to vary widely across security assistance programs and authorities in terms of sufficiency and quality.

Fundamentally, security sector assistance will often take years to yield results, whereas incidents of civilian harm can occur and escalate quickly. Policymakers and practitioners should aim to calibrate security partnerships to reflect these risks and ensure that U.S. expectations of the partner do not exceed its capabilities.

Norms of Conduct

In the twenty-first century, the means and methods of warfare continue to evolve. Operational adaptability of the United States and its partner forces is of paramount importance, particularly where adversaries employ tactics that place civilians at risk.

The United States risks being perceived as colluding to allow harmful behavior if it seeks to deconflict or de-escalate during an operation (e.g., establishing de-escalation zones in Syria that in practice allow the Syrian regime and Russian and Iranian-backed agents to fortify their positions and target civilians elsewhere). In these complex environments, the United States and its partners may be inclined to pursue a “lowest common denominator” approach to transparency and accountability to avoid conflicting (and potentially embarrassing) messaging. However, this approach could simply postpone instituting effective practices to heighten operational adaptability and mitigate civilian harm.


Key Considerations and Models of Partnership


In any security partnership, U.S. policymakers and practitioners should seek to address key considerations to mitigate the potential for civilian harm by partner security forces:

  • Defining U.S. responsibilities to prevent, monitor, and account for harm incurred during its partner’s operations;
  • Assessing the risks involved, to include a comprehensive analysis of security governance, accountability, and oversight mechanisms available to regulate security force conduct;
  • Understanding national and local politics and incentives playing out in the conflict when assessing risk to civilians;
  • Calibrating the model of operational oversight that most effectively shapes the conduct of partner operations;
  • Encouraging partners to incorporate practical measures to mitigate and investigate instances of civilian harm in their operations;
  • Cultivating open channels of communication with civil society that can be used to communicate intentions, report incidents, prevent problems, and address civilian protection-related concerns;
  • Ensuring that plans and efforts to collect information or intelligence from civilians to counter or respond to physical threats from other actors will not unintentionally expose civilians to greater risk; and
  • Identifying conditions or partner choices that would prompt limits on or suspension of the partnership.

Furthermore, the United States lacks a common government-wide policy or doctrinal definition for characterizing different security partnerships in implementing a “by, with, and through” approach. The following typology could be used to distinguish a range of security partnerships:6

1) Support Operations

Providing materiel, intelligence, training, and other forms of support without directly undertaking airstrikes or ground operations.

Example: Refueling support, advice on targeting, intelligence, arms sales, and training for Saudi Arabia in its Yemen operations

Specific Considerations for the United States:

  • Validating the integrity of any information used in lethal operations and that the intelligence is used for its intended operational purpose; and
  • Adapting the provision of materiel to ensure proper use, including customized approaches to assistance and advising, and conducting robust presale risk assessment of material and training assistance and end use monitoring.

The U.S.-Saudi Security Partnership
The United States maintains counterterrorism objectives in Yemen vis-à-vis al Qaeda in the Arabian Peninsula and Iranian support for the Houthi rebels. However, the Saudi-led intervention has prompted concerns about civilian casualties and damage to civilian infrastructure, the de facto blockade on essential goods entering Yemen, and access for humanitarian operations in the broader context of Yemen’s civil war. The U.S.-Saudi partnership falls under a model that has not yielded concrete results in terms of mitigating civilian harm in the context of protracted hostilities, carrying implications for the effectiveness of operations in meeting security objectives.

2) Advise, Assist, and Accompany Missions

Limited presence of U.S. forces on the ground and, with the consent of the partner, provide intelligence, training, advice, and other forms of technical assistance to partners who conduct operations.

Examples: Joint Special Operations Task Force–Philippines; Joint Special Operations Task Force–Juniper Shield (Niger)

Specific Considerations for the United States:

  • Adapting training, advice, and other forms of technical assistance on a continuous basis to ensure the partner’s ability and capacity to mitigate civilian harm as operations evolve;
  • Understanding past conduct and perceptions of partner security forces and the potential risks involved with U.S. forces’ association with them;
  • Preparing for the contingency that U.S. forces become engaged in the use of force in self-defense and, as a result, the potential for civilian harm by U.S. forces; and
  • Ensuring transparency and adequate public consultation on the nature of U.S. activities.

U.S. Participation in Advise, Assist, and Accompany Missions
U.S. advise, assist, and accompany operations may involve rules of engagement and assets that enable the use of force for self-defense, but do not typically involve direct U.S. participation in hostilities, and may specifically proscribe direct participation by law or formal agreement. Due to the lack of direct U.S. involvement in hostilities, encouraging partners to ensure the protection of civilians can be more difficult from the U.S. perspective.

3) Regional Platforms and Configurations

Cases in which U.S. forces may not be directly involved but play a role in shaping the policy and planning for regional security collectives and configurations, provide funding, or provide bilateral training or other forms of support to parties.

Example : G5-Sahel

Specific Considerations for the United States:

  • Leveraging the U.S. role as a funder or source of influence to ensure proper structures of compliance and accountability;
  • Encouraging these platforms to mitigate harm to civilians in their operations and adopting specific and interoperable policies and practices on civilian harm mitigation; and
  • Promoting partner command climate that emphasizes the importance of preventing and accounting for civilian harm.

French and Malian soldiers speak during counterterrorism operations in the Sahel — Philippe Desmazes/AFP/Getty Images


4) Partnered Operations

Cases in which the United States is directly involved in planning and actively involved in hostilities or the use of force in addition to possibly advising or assisting on the ground.

Example: Providing air support to Iraqi national security forces in operations against ISIS

Specific Considerations for the United States:

  • Including prevention, investigation, and mitigation of civilian harm mechanisms in tactics, techniques, and procedures for U.S. forces operating with partner forces (e.g., U.S. Army Security Force Assistance Brigade in Afghanistan and beyond);
  • Mitigating risks associated with the use of force when depending on partner intelligence; and
  • Anticipating possible risks to civilians associated with combining U.S./coalition air operations and partner ground force operations and tailoring preventative, investigative, and mitigation measures accordingly.

U.S.-Iraqi Partnered Operations
In its efforts to wrest control of its territory back from the Islamic State, the Iraqi government receives support from a U.S.-led coalition of states called the Combined Joint Task Force Operation Inherent Resolve (CJTF-OIR). The coalition conducted air strikes as well as trained and advised Iraqi Security Forces (ISF) in operations to retake the city of Mosul. Planning for the Mosul operation was led by ISF with coalition participation through the Combined Joint Operations Center; coalition support entailed building ISF capacity through training, providing intelligence, and fire and technical support. Ground fighting around and within Mosul was led by ISF, but Iraqi forces could call in U.S. and coalition air support. While U.S. Central Command continues to publish monthly reports on civilian casualties attributed to its operations in Iraq, no such public reporting or assessments of civilian harm attributed to ISF exist within Iraqi government agencies.7

5) Partnered Operations with Armed Nonstate Actors

At times, the United States may support local armed groups that are not a part of a national military or security force when pursuing goals in fragmented states and civil wars. Some nonstate armed groups have a relatively long history, clear political objectives, and strong internal cohesion with a functioning command and control. Others are newly formed, with fluid structures and evolving political objectives.

Examples: Support to and joint operations with the Syrian Democratic Forces and Iraqi Kurdish peshmerga

Specific Considerations for the United States:

  • Evaluating how to factor the post-conflict future of the armed group into U.S. support—for example, should the United States support disarmament and demobilization of nonstate armed groups following the cessation of hostilities or, alternatively, their integration into national armed forces?




Deeper research and analysis surrounding these issues is warranted.8 Given increasing reliance on the “by, with, and through” approach by the U.S. military, policymakers and practitioners should take the following near-term actions to mitigate the potential for civilian harm:

  • Develop common and consistent definitions for the range of “by, with, and through” security partnerships in U.S. national security and defense policy;
  • Establish standard protocols for identifying and mitigating risks involved with partnership (e.g., overly permissive partner nation rules of engagement, lack of military investigative capacity, past conduct issues);
  • Implement and publicly communicate a U.S. Department of Defense directive to guide policy, procedures, and process for the protection of civilians, as well as public investigations and reporting on civilian harm, in partnered operations;
  • Early in the partner relationship, establish standard protocols for attribution, investigation, and remedy for civilian harm in partnered operations;
  • Design security partnerships to emphasize the protection of civilians as an essential component of military effectiveness through professionalization, institution building, scenario-based exercises, operational training, and advising;
  • Acknowledge there will be perennial tensions in U.S. policy surrounding this issue, including:
    • Be clear-eyed about when operational imperatives will override concerns about civilian harm, including a rigorous cost-benefit analysis of “by, with, and through” partnership;
    • Recognize that instances of civilian harm require immediate attention for effective mitigation of repeated patterns of harm as well as response to harm that occurs, while the best solutions and mitigation often must occur “upstream” in the relationship via institution building, scenario-based exercises, and advising; and
    • Be cognizant of placing high expectations on partners related to their conduct and the geopolitical reality of entering into partnerships based on near-term opportunities and short timeframes, which may necessitate quick agreements and potentially little preplanning or training.
  • Increase engagement with the public, media, and civil society to identify specific means by which the United States can work with partners to mitigate civilian harm, investigate reports of misconduct by U.S. and partner forces, and hold personnel accountable.

CSIS Briefs are produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).

© 2018 by the Center for Strategic and International Studies. All rights reserved.

[1] Melissa Dalton is a senior fellow and deputy director of the International Security Program and the director of the Cooperative Defense Project at the Center for Strategic and International Studies (CSIS). Daniel Mahanty is the director of the U.S. Program at the Center for Civilians in Conflict (CIVIC). Jenny McAvoy is the director of protection in the Humanitarian Policy and Practice Team at InterAction. Hijab Shah is a research associate with the International Security Program at CSIS. Julie Snyder is a research and advocacy associate with the U.S. Program at CIVIC. Kelsey Hampton is a policy coordinator for protection in the Humanitarian Policy and Practice Team at InterAction. This paper is a part of a collaboration by the CSIS Cooperative Defense Program, CIVIC, and InterAction on the protection of civilians in partner operations.

[2] The legal obligation to distinguish combatants from noncombatants in war is fundamental to the laws of armed conflict (along with proportionality and precaution) to ensure that civilians and civilian objects are spared from the effects of conflict. In practice, adhering to the principle of distinction can be complicated by a range of factors, especially in armed conflicts of a noninternational character (civil wars) or when armed groups do not distinguish themselves through uniforms or insignia or operate among civilians or in populated areas.

[3] Minimizing civilian harm is a stated priority for the U.S. military, but it can pose challenges at the operational and tactical levels and can take on additional complexity in partnered operations. In addition to the harm suffered by the victim, civilian casualties and damage to civilian infrastructure may have a range of negative effects on the party that caused them. If left unattended, civilian harm can affect morale and discipline of the force. Civilian casualties can erode support for a military campaign, galvanize passive or active support for the adversary, and even result in a lack of adversary compliance with international humanitarian law, imposing operational challenges to the United States and its partners. Civilian harm may also lead to perceptions of ineptitude or a lack of credibility. The United States and its partners might better mitigate civilian harm and improve leadership credibility through the following measures: a more critical examination of targeting intelligence; incorporation of nongovernmental organization and civil society reporting into operational planning, systematic assessments and analysis of civilian harm and the effects of military operations on civilian infrastructure; flexibility in updating and adapting operational plans at various stages of the planning process; responding in a timely manner to reports of civilian harm; increasing transparency on investigation processes; providing clarity on responsibility for civilian harm; and possibly providing compensation for incidents of civilian harm. See Melissa Dalton and Joseph Federici, “Operational Security, Accountability, and Civilian Casualties.” CSIS Critical Questions, November 30, 2017,

[4] There are a range of legal aspects to partnered operations from applicable international humanitarian law, state responsibility in coalitions, international criminal responsibility, and U.S. national legislation that seeks to prevent U.S. security assistance to forces that commit human rights violations. These legal issues would lend additional insight to future research but are beyond the scope of this policy brief.

[5] U.S. policymakers and practitioners should set realistic expectations and examine best practices and innovations that consider who does the training, what is known about how to best affect partner behavior, how best to customize the full range of assistance to local requirements, what tools are most effective, when they make the biggest difference, and how to monitor and evaluate performance over time.

[6] In some cases, certain models of partnership may have their own specific considerations for mitigating civilian harm. It is also important to recognize that some partnerships might transition from one model to another over time, depending on U.S. and partner requirements.

[7] See Center for Civilians in Conflict and InterAction, “Protection of Civilians in Mosul: Identifying Lessons for Contingency Planning,” October 17, 2017,

[8] The U.S. policy and practitioner community should conduct research and develop frameworks and recommendations surrounding: U.S. policy opportunities, challenges, and gaps in each model of security partnership, particularly as it relates to congressional interest in the issue; partner motivations and incentives to play a greater role in civilian harm mitigation, incorporating these findings into training, advising, exercises, and institution-building programs; effectiveness of existing legal structures (U.S. and international) in mitigating civilian harm, and providing recommendations for strengthening them; calibration of operational oversight with partner-led interventions to avoid moral hazards or deeper interventions into local conflicts; and developing an “exit strategy” or “off ramps” for the United States in instances where the partners grossly violate the humanitarian tenets of U.S. security assistance policy or are unwilling to mitigate civilian harm in conflict.

Daniel Mahanty

Daniel Mahanty

Former Senior Associate (Non-resident), Humanitarian Agenda

Julie Snyder

Jenny McAvoy

Director of Protection, Humanitarian Policy and Practice Team, InterAction

Kelsey Hampton

Policy Coordinator, Protection, Humanitarian Policy and Practice Team, InterAction

Hijab Shah

Melissa Dalton