Not a week had passed since the Sudanese Armed Forces expelled the Rapid Support Forces from central Khartoum in May 2025, when social media began to fill with posts from displaced families planning their return to the capital. These expressions were quickly met with statements from international humanitarian actors emphasizing that conditions in Khartoum did not meet minimum standards for safe, dignified, or sustainable return. Displaced Sudanese are aware that returns may still be unsafe, just as Syrians were in late 2024 following the ouster of Bashar al-Assad, and Iraqis after the territorial defeat of the Islamic State of Iraq and the Levant (ISIL) in 2017.
But in these three cases, those who saw a “good enough” opportunity to return and were willing to take the risk, often did so relatively rapidly. This is why warnings against premature returns must now be accompanied by a broader conversation on what is needed to make areas of return genuinely viable: What matters the most for those already returning, and for those still waiting for the bare minimum opportunity to follow? What investments can most effectively restore a sense of stability and normalcy? What should be prioritized and by whom?
Overarching drivers of return: housing and protection
We sought to better define what bare minimum conditions are across these three countries using microdata from the International Organization for Migration’s Displacement Tracking Matrix (DTM). In each case, IOM DTM tracked displacement, returns, and living conditions across thousands of conflict-affected villages, towns, and neighborhoods. The indicators we developed covered similar themes, spanning both material and social dimensions of daily life, tailored to each specific conflict context. By quantitatively modeling the differences between locations, our aim was to understand why returns occur in some areas and not in others. Our central argument was that some indicators weigh more heavily in returnees’ decision-making than others. This became evident through statistical modeling, where findings across all three contexts revealed a consistent narrative: people prioritized housing conditions and protection. Where individuals can find shelter and some assurance that they will not be targeted nor face retaliation, they are far more likely to return. Other factors often prioritized first as essential needs in response frameworks, such as access to basic services, economic activity, or functioning markets, remain important for long-term sustainability and quality of life, but are generally viewed as secondary, or as challenges that can be managed upon return.
We first pioneered this type of modeling on return dynamics in Iraq. There, the end of the three-year conflict with ISIL in 2017 triggered a surge in displaced people attempting to return to their areas of origin. IOM DTM had recorded 1.5 million returnees by January 2017. This quickly increased to 3.4 million a year later and 4.2 million after two years. Many areas regained their pre-conflict populations, but by early 2019 roughly 20% of the locations monitored still saw either minimal returns or no returns at all. These locations fared worst, offering limited prospects for wellbeing for those seeking to return. Fitting a model that compared the living conditions in locations that received returnees with those that did not, allowed us to empirically identify what made a location conducive to returns.
Residential destruction emerged as the most decisive factor for Iraqi returnees. Locations where at least half of the housing stock was destroyed were 15 times less likely to see returns than those with no destruction. Access to employment followed as another significant driver. Beyond these two material indicators, however, a second tier of factors related to the socio-political landscape of Iraq’s conflict-affected areas proved influential. These included the need for community reconciliation, the presence of a multiplicity of security actors (including hybrid militia groups that fought ISIL, wield significant power, but lack the trust of local population), and the enforcement of return bans on certain families, which themselves connected to the absence of meaningful reconciliation and justice processes. Other indicators included in the modeling such as access to basic services, public utilities, functioning of markets, or agricultural activity, while arguably essential for long-term recovery, had comparatively less impact on the immediate decision to return. In many cases, people returned despite the absence of these.
These findings were important because they provided a strong evidence base for reframing the post-conflict narrative in Iraq, from one that prioritized infrastructure repair as a driver of returns to one that proposed a diversified approach for resolving displacement, looking beyond return movements alone to consider the root causes of the conflict. In short, better integrating humanitarian, development, and peacebuilding actions in the post-conflict landscape. Micro-data over time reveal that material living conditions steadily improved in return locations as had bottom-up community reconciliation, while return-related security concerns, particularly involving hybrid militia groups, worsened. This is in part because security assistance, operating in parallel but separately from humanitarian, development, and peacebuilding efforts, focused primarily on counterterrorism, priorities of Iraqi authorities and their international partners, rather than the full remit of civilians’ protection needs. This has left around a million people still displaced without a choice of resolution, carrying with it lasting rights implications.
We carried out similar modeling in Syria in 2020, during a period of stalemate in the civil war and ongoing territorial fragmentation. The country had already experienced waves of forced displacement, demographic reengineering, and negotiated population transfers. While regional patterns varied, data indicated that four in ten individuals in conflict-affected areas had been displaced. Our analysis sought to understand why some communities managed to retain their populations while others did not, that is, what factors drove depopulation.
We found a correlation between poor living conditions and high levels of population loss. As in the Iraq case, the most critical factor in Syria affecting a location’s ability to retain its population pertained to housing. While the destruction of homes was a primary driver, two other related factors also stood out: the presence of unexploded ordnance in residential areas, and the illegal occupation of housing. Both were salient factors that explained protracted depopulation, with the latter occurring in areas that had unaddressed historical antecedents of forced displacement, demographic change, and land and property seizures.
Safety and security-related factors followed closely. Restrictions on movement, sexual harassment of women and girls, and forced recruitment of adult men all came to the fore as second-tier drivers of displacement. Concerns over arbitrary arrest and the conduct of state and non-state security actors operating in a given location were also strongly linked to people’s decision to flee. As was the concern over retaliation should others come back. While these security risks and threats of violence were widespread and persistent across the country, and again had unaddressed historical roots, what shaped whether people stayed was not their presence, but the intensity in which they were experienced. Widespread poor service provision and deteriorating economic conditions in Syria, on the other hand, appeared to have less influence on patterns of mobility. All of these factors are now compounding upon each other since 2024 as Syrians have started to return (voluntarily or not) or consider doing so.
Finally, the replication of this modeling in Sudan took place in 2022, prior to the outbreak of the current conflict. It focused on the Darfur and Kordofan regions, tracking patterns of displacement and return linked to the conflict that began in 2003 and was never fully resolved. The aim here was to contribute evidence to support ongoing efforts at broadening the response toward a more integrated humanitarian, development, and peacebuilding approach.
Once again, the most decisive factor that had allowed some people to return was safety over other material considerations. Although conflict intensity reduced since the signing of the Juba Peace Agreement in 2020, many displaced people feared the continued and unaddressed bouts of extreme violence against civilians. In locations where returns remained absent, fears centered on tribal violence, sexual and gender-based violence, and a general lack of law enforcement. A secondary set of obstacles included food scarcity, residential destruction, and the disruption of pre-existing customary dispute resolution mechanisms (notably, there was a significant distrust in state-sponsored mechanisms). This finds resonance in Darfuri demands for justice ahead of renewed reconciliation pushes, right before the entire country plunged into conflict – and in the dynamics those returning to Khartoum are finding as well. Among secondary factors, food availability stood out as the only life-saving factor among the top barriers to return. This fits in contexts like Darfur and Kordofan, where the risk of deprivation is particularly acute compared to the other countries assessed.
These empirical findings stand in stark contrast to cynical, xenophobic narratives and ensuing policies that posit return at all costs, simply because conflict ends. Rather, they highlight what and how much is needed for such movements to be truly safe and sustainable.
Against standard practice
From a normative perspective, all the dimensions considered are undoubtedly important to people’s lives, especially in the context of conflict. However, the consistent lack of a strong causal relation between lifesaving aid or essential needs and actual patterns of human mobility indicates that displaced populations prioritize differently. Housing, safety, and social stability consistently emerged as more existential concerns for displaced people considering returns than, for instance, having the lights on.
The incongruence lies in that many of the aspects that seem to matter for return decision-making are often overlooked in early-stage crisis response or are poorly integrated into broader response frameworks. These responses prioritize the immediate, visible needs of displaced populations, aiming to assist people on an individual basis where they are now, rather than the needs of the communities where they are from. This approach, while often necessary in an initial phase, can further protract displacement by failing to address root causes that prevent return. Moreover, tackling more structural conditions like housing and security is frequently deemed too political or beyond the remit of humanitarian actors. These efforts are instead frequently assumed by local actors or de facto authorities who are well-placed but under-resourced and often exposed to risk. The existence of domestic IDP policies could help to mitigate these discrepancies, setting the direction for a wider response by surfacing issues that are relevant to affected populations. How well such state policies align with the priorities and rights of the displaced are, however, dependent on their levels of formality and liberality, which vary considerably by context.
New frameworks like the humanitarian-development-peace triple nexus and related debates around localization and aid neutrality have also begun shifting practices to address these constraints. Progress is slowed by conceptual confusion over components of the triple nexus, the retrenchment of power imbalances in localization paradigms, and an unwillingness and inability to politically end crises. As such, the ongoing mismatch between what matters the most to affected populations and what responses systems are structured to deliver inherently takes the impact out of them. This has left the sector open to mounting scrutiny and existential threat from increasingly bad faith actors who question their long-term relevance and effectiveness, all while global displacement numbers continue to rise and armed conflict spreads. The United Nations’ response to reductions in foreign aid spending, shrinking even before last year’s dismantling of USAID, seems to imply a consensus of doing the same (or less) with less resources.
Everything, all at once
The findings on Iraq, Syria, and Sudan discussed here, however, call for greater imagination: an unabashedly maximalist vision to work from, not a shrunk down status quo. The decision-making and return movements of displaced people reinforce the need to consolidate a transversal approach, fully embedding the “peace” dimension (with all its implications) to what has too often been limited to survival-focused aid or constrained development programs.
This requires a willingness to engage politically in crises: creatively and in coordination with wider swaths of stakeholders, following the lead of displaced and conflict-affected populations rather than focusing on political expediency or geopolitical interests alone. This holds especially true for hard security elements of peace linked to protection of civilians and security configurations that are of salience to affected communities, which often cannot be resolved at the community-level alone.
Of critical importance is the recognition that fostering peace, restoring social fabrics, and involving actors more ready to engage with political and security matters must start early in any response, centered on the wellbeing, rights, and aspirations of impacted communities. Only by aligning with what conflict-affected populations identify as most important is it possible to meaningfully address displacement, reduce protracted crises, and restore a sense of agency and hope. Achieving this requires investing in everything across the board: not in sequence, but simultaneously.
Roger Guiu and Nadia Siddiqui co-direct Social Inquiry, a research organization whose work has aimed to improve the impact and effectiveness of public policies and interventions to address the root causes of conflict in fragile societies. We do this through cross-disciplinary applied research focused on governance, civil trust, and inequality. This has included developing innovative studies exploring what helps or hinders the resolution of internal displacement, examining material, spatial, and socio-political conditions and individual, collective, and structural factors.
This topical paper is part of the special series on ‘Internal Displacement in a Changing World Order’, led by the Internal Displacement Research Programme at the RLI. The experts contributing to this series assess how rapid shifts in contemporary politics, plummeting levels of humanitarian aid and escalating global crises are impacting displacement-affected communities. The series ties into a recently launched 45-chapter “Handbook of Internal Displacement” (2026) that comprehensively addresses this issue.
KEYWORDS: Internal Displacement, IDPs, decision-making, solutions.
Selected bibliography
Bradley, M. (2018). Durable Solutions and the Right of Return for IDPs: Evolving Interpretations, International Journal of Refugee Law 30 (2): 218-242. https://doi.org/10.1093/ijrl/eey021
Nguya, G. & Siddiqui, N. (2020). Triple Nexus Implementation and Implications for Durable Solutions for Internal Displacement: On Paper and in Practice, Refugee Survey Quarterly 39, 466–480. https://doi.org/10.1093/rsq/hdaa018
Siddiqui, N. & Guiu, R. (2024). HDP Nexus Outcomes and the Integration of Peace Elements: The Case of Iraq after the ISIL Conflict, The International Spectator 59 (3): 118-137. https://doi.org/10.1080/03932729.2024.2343285