Discussion Summary

The event brought together policymakers, practitioners, and researchers to examine how migration decisions can be influenced at different stages: before departure, along migration routes, and at points of return or onward legal pathways. Across the day, the discussion moved from individual-level interventions to broader systems-level questions, with a consistent emphasis on evidence, adaptability and responding to rapidly changing migration dynamics.

Key takeaways

The greatest return on investment comes from early, upstream interventions. As journeys progress, it becomes harder for migrants to access – and believe in – alternatives. Later in the journey, chances of encountering border security apparatus and asylum systems is greater, which comes at a greater cost to migrants and governments alike.

Integrate information with counselling, referrals and practical support. Information alone rarely changes behaviour. The event heard multiple approaches to this involving remote-first methods, government partnerships, physical focal points, and more. Reliable measurement of behaviour change is possible using longitudinal and quasi-experimental approaches.

Trust is context-specific and shifts along the route. Diaspora, peers, digital channels and physical centres each play different roles. Connect the system, not just the service. Interventions work better when delivery channels, referral pathways and service providers are aligned rather than operating in isolation.

Make alternatives feel real enough to choose. Regular pathways, protection options, and return and reintegration support influence behaviour only when they are visible, realistic and attainable.

Discussion insights

Discussions were conducted under the Chatham House rule, therefore the following points are summarised by theme and not attributed to speakers.

Opportunities abound to improve upstream intervention

Migration programming takes place in a context of increasing mobility, where decisions are complex, highly contextual, and constantly evolving. Simple explanations or single interventions rarely work, and information alone is often insufficient to change behaviour when filtered through distrust, peer influence, and lived experience.

Evaluation of Seefar’s Missing Migration Links (MML) pilot in Afghanistan and Ethiopia suggested that of 2,500 referred potential migrants, nearly 20% in Afghanistan and 13% in Ethiopia abandoned their migration plans as a result.

AI-supported identification, such as MML’s use of Pintone.ai, can help programmes better target those most at risk. Economic motives emerged as a key factor in identifying individuals at heightened risk of irregular migration, reinforcing the need for interventions that combine information with training or job opportunities, and raising the question of whether similar approaches could identify the most vulnerable for humanitarian support.

Netherlands-supported programming in Morocco found that nearly 70% of people under 30 wanted to leave.

Gamified approaches reach scale for a low cost, with over 25,163 students participating in teacher-led escape games that are now part of the national curriculum.

Alternatives must feel attainable if they are to influence migration at scale

Alternatives such as legal migration pathways or assisted return only shape behaviour when they feel genuinely attainable and within people’s control. Effective return and reintegration depend on psychosocial readiness and sustained, layered support over time so that individuals feel ownership over their decisions.

The mere perception that legal migration pathways are available can influence irregular migration as strongly as access to visas themselves. For instance, after the United States agreed in 1995 to offer 20,000 legal places to Cubans, interceptions at sea dropped dramatically: from 34,000 (January–August 1994) to just 525 the following year. Similarly, irregular encounters among nationalities covered by Safe Mobility Offices fell by 91% within two years of the programme’s introduction. In Germany, asylum applications from the Western Balkans decreased from 120,000 to 34,000 after work visa quotas were doubled.

Along the route: trust, influence, and opportunities for cooperation

Route-based communications and referral programmes show that their effectiveness depends on trust, which is not automatic. Trust depends heavily on where people are in their journey, the credibility of the information source, and the pressures they are facing at the time.

There are no universal rules for engagement; the question is rather: when, how and for whom does communication actually work? Interventions have therefore taken multiple forms, from physical access points such as Migrant Resource Centres offering counselling, referrals, and job fairs, to long-term engagement with diaspora communities, and large-scale digital or telephone-based counselling reaching tens of thousands of identified irregular migrants.

Discussion highlighted the growing importance of digital outreach in transit contexts, where these channels are often the primary way to reach people on the move. However, once individuals are already in transit, there may be a greater need to pair information with practical support, navigation, and follow-up. At the same time, digital-first does not mean digital-only; the most effective programmes combine online engagement with in-person and other approaches.

Assisted return is a prevention tool

Ultimately, counselling and awareness-raising only have impact when they connect people to services and pathways that are accessible, credible, and sustained. This includes further exploring the role of assisted voluntary return in transit countries. Without active follow-up and clear referral mechanisms, even well-designed interventions risk becoming administrative exercises rather than meaningful pathways for individuals.

Measuring the contribution of interventions to reduced migration is possible

It is notoriously difficult to measure behaviour change and its contribution to migration policy objectives. However, data from projects shows longitudinal and quasi-experimental approaches can balance rigour with practical feasibility. For example, under the AMIF-funded Pathways, a longitudinal study found that 70% of those in counselling abandoned their plans, while only 13% still intended to migrate after referrals. In addition, realist evaluations can help implementers understand what works in a specific context without the need to reach conclusions that apply globally.

Next steps

Over the next year, irregular migration prevention and adjacent activities will grow in number, including the launch of the Global Alliance Rapid Response Initiative, AMIF-funded communications activities, bilateral migration and development programming funded by European interior and foreign ministries, and initiatives supported by private donors. Seefar will seek opportunities to formally and informally strengthen coordination between them and develop overarching theories of change for how programming can further develop in a way that is ethical and effective.

Thanks to

Contributions from:
European Commission
Migrationsverket
ICMPD
DRC
MPI

For additional materials, including presentations on pilot interventions and programme insights, as well as video contributions, please contact [email protected]