Stay up-to-date with what we are doing

Blog Series: Part 1

The Good and Bad of Anticipating Migration

Posted on 9th of October 2023 by Stefaan Verhulst, Sara Marcucci

The Good and Bad of Anticipating Migration
The Good and Bad of Anticipating Migration

This blog is the first in a series that will be published weekly, dedicated to exploring innovative anticipatory methods for migration policy. Over the coming weeks, we will delve into various aspects of these methods, delving into their value, challenges, taxonomy, and practical applications. 

This first blog serves as an exploration of the value proposition and challenges inherent in innovative anticipatory methods for migration policy. We delve into the various reasons why these methods hold promise for informing more resilient, and proactive migration policies. These reasons include evidence-based policy development, enabling policymakers to ground their decisions in empirical evidence and future projections. Decision-takers, users, and practitioners can benefit from anticipatory methods for policy evaluation and adaptation, resource allocation, the identification of root causes, and the facilitation of humanitarian aid through early warning systems. However, it's vital to acknowledge the challenges associated with the adoption and implementation of these methods, ranging from conceptual concerns such as fossilization, unfalsifiability, and the legitimacy of preemptive intervention, to practical issues like interdisciplinary collaboration, data availability and quality, capacity building, and stakeholder engagement. As we navigate through these complexities, we aim to shed light on the potential and limitations of anticipatory methods in the context of migration policy, setting the stage for deeper explorations in the coming blogs of this series.

Why Anticipate Migration? Five Reasons Why Anticipatory Methods Are Valuable For Migration Policy

Human migration is a global phenomenon. According to the IOM’s World Migration Report 2022, there were around 281 million international migrants in the world in 2020, which equates to 3.6 per cent of the global population. While certain migration drivers, like economic or environmental factors, have traditionally been analyzed independently, there is a growing recognition that migration does not result from a single factor. Instead, it emerges from intricate combinations of multiple, interrelated, and interdependent factors.

As illustrated by the International Migration Institute, these diverse drivers encompass demographic factors, such as changes in population size and composition, economic disparities that include income levels and employment opportunities, environmental factors like climate change and natural disasters, and human development indicators that define living standards and access to education and healthcare. Additionally, individual motivations and resources, along with socio-cultural norms and networks, play pivotal roles in migration choices. Political and institutional policies, regulations, and security conditions within and between countries also significantly influence migratory flows. Furthermore, supranational forces, such as global economic systems, transnational ties, regional integration, and geopolitical shifts, exert their influence on migration dynamics. Together, these multifaceted drivers paint a comprehensive picture of the intricate forces shaping human mobility in today's interconnected world. 

The 2015-2016 refugee crisis in Europe, which saw the arrival of hundreds of thousands of refugees and migrants, underscored the pressing demand for enhanced levels of preparation and anticipatory strategies in the realm of migration policy. Indeed, many European countries struggled to process, accommodate, and care for asylum seekers, making the need for proactive anticipation and readiness measures ever more evident. 

Amidst the dynamic and intricate landscape of migration, anticipatory methods indeed emerge as valuable tools that may offer guidance to policymakers. They can provide a framework to better understand the complexities of contemporary migration patterns, enabling governments and international organizations to transition from a reactive to a more proactive approach. These methods encompass a range of techniques, from predictive models that utilize current data to forecast short-term migration events, to foresight methods that allow policymakers to explore and influence potential migration scenarios

In parallel, following the advancement of technology and data-based tools, the demand for migration data has increased, leading to experimentation with numerous, non-traditional sources, including mobile phone, banking, social media, and other forms of big data. These rich and diverse data sources, coupled with anticipatory methods, not only have the potential of enhancing our ability to mitigate the uncertainties surrounding migration but also of offering a pathway to develop more precise, responsive, and forward-thinking policies.  

While these methods do not claim to eliminate uncertainty entirely, they seem to offer a means to develop more precise, adaptable, and forward-thinking policies. In an era where the movement of people continues to evolve, anticipation can offer a multitude of potential benefits, including enabling evidence-based policy development, facilitating policy evaluation and adaptation, supporting strategic resource allocation, identifying root causes, mitigating immediate risks, and fostering international cooperation and governance. Below, we delve into each of the aforementioned reasons, exploring how anticipatory methods can contribute to more informed, resilient, and proactive migration policies.


Figure 1: Potential Benefits of Innovative Anticipatory Methods

  • Proactive Policy Development: Anticipatory methods can enable more proactive policy development. For instance, the EU Migration Preparedness and Crisis Management Mechanism (Blueprint) Network is an operational framework for monitoring and anticipating migration flows and migration situations, building resilience, and organizing a coordinated response to migration crises. The network seeks to exchange information on migration trends and ensure continuity with what the Council's emergency mechanism currently does. Indeed, the purpose is to develop an early warning and forecasting system to anticipate migration flows and trends, being complementary to other EU crisis management mechanisms, such as the EU Civil Protection Mechanism and the Integrated Political Crisis Response (IPCR). The system uses a range of data sources, including satellite imagery, social media, and other sources of information, to identify potential migration hotspots and anticipate future migration flows, thereby aiming to enable enhanced preparedness, proactive governance, and timely responses.
  • Policy Evaluation and Adaptation: In an uncertain world, the ability to adapt is crucial. Anticipatory mechanisms can set the stage for policies that are more flexible and adaptable, designed with the understanding that conditions may change. For instance, anticipatory methods developing counterfactual inference using past time series data with plausible migration policies scenarios can be used to analyze the effect of search-and-rescue missions on the number of crossing attempts of migrants across the Mediterranean, allowing policy makers to evaluate existing policies and adapt them based on their effectiveness. 
  • Resource Allocation: Through the analysis of migration data and by anticipating future migration patterns, policymakers can pre-position resources, establish contingency plans, and implement proactive measures to respond effectively. This includes allocating resources such as housing, healthcare, education, and social services in a targeted and efficient manner, ensuring efficient and equitable distribution of support.  For instance, the Anticipatory Pillar of the Disaster Response Emergency Fund (DREF) is an initiative by the International Federation of Red Cross and Red Crescent Societies (IFRC) that enables Red Cross and Red Crescent Societies to take early action before disasters strike. The Anticipatory Pillar uses a forecast-based financing approach based on meteorological forecasts and risk analysis that helps direct funds automatically, when pre-defined forecast thresholds or 'triggers' are met. 
  • Identifying Root Causes: By analyzing data and projecting future scenarios, policymakers can identify underlying factors that drive migration. Indeed, through the analysis of both historical and present data, decision makers can evaluate a variety of drivers and scenarios, including poverty and unemployment rates, access to education and healthcare, as well as conflict and environmental degradation rates. Initiatives like the one undertaken by the UNHCR Innovation Service employ foresight methodology including signal mapping, causal layered analysis, and scenario archetypes to surface new stories and worlds to address increased arrivals of asylum seekers and related challenges. The report aims to identify overlooked aspects of the refugee experience and the components of a given crisis, exploring the current and possible narratives associated with a refugee crisis to surface new root-cause insights. When policymakers and organizations identify the root causes of migration through data analysis and scenario projection, they can develop comprehensive strategies that hope to tackle root causes, focusing on long-term solutions to reduce forced migration and promote sustainable development. 
  • Humanitarian Aid: Anticipatory methods can facilitate the establishment of early warning systems and coordinated response mechanisms that span across multiple countries, thus promoting humanitarian aid through cross-border cooperation, harmonization of migration policies, and the development of common goals and standards. For instance, the Famine Early Warning System Network (FEWS NET) is an initiative that uses anticipatory methods to provide projections of food security outcomes for the coming eight months. The FEWS NET analysts adopt a standardized eight-step process to assess the current food security situation in areas of concern, make assumptions about the future, and consider how those assumptions might affect food and income sources for poor households. More specifically, the initiative uses remote sensing satellite imagery to monitor and forecast climatic conditions in the world’s most food-insecure regions, and analyzes drivers of acute food insecurity, including economic, social, natural and political factors. Overall, FEWS NET's anticipatory methods help to establish early warning systems and coordinated response mechanisms that span across multiple countries, promoting humanitarian aid through cross-border cooperation, harmonization of migration policies, and the development of common goals.

It is important to note that despite the uptick in human mobility and technological advances to gather and (re)use data on and from migration patterns, there still exists a notable challenge in effectively foreseeing, anticipating, and preparing for large movements of migrants, creating bureaucratic burden, inefficient resource use, and reactive support for people on the move. The following section will explore some of the challenges related to using anticipatory methods for migration policy. 

Challenges Related to Using Anticipatory Methods

The challenges associated with predictions and anticipatory methods in the context of migration policy and governance are multifaceted. All in all, it seems possible to divide them into two main categories, namely (a) the conceptual challenges of adopting anticipatory techniques, which span from algorithmic bias to uncertainty, from the legitimacy of preemptive intervention to group privacy, from self-fulfilling prophecies to the politicization of predictions, and (b) the implementation challenges of mitigating those challenges, successfully employing and executing the methods, including issues of collaboration, data availability and quality, capacity building, and stakeholders engagement. 


Figure 2: Conceptual and Implementation Challenges of Innovative Anticipatory Methods

Conceptual challenges

  • Bias: One key challenge is the issue of fossilization, whereby algorithmic predictions reinforce patterns in data about the past, potentially perpetuating pre-existing bias and inequality. This can lead to the entrenchment of existing marginalizing patterns in the context of migration, and hinder efforts to address systemic issues. One example is the use of predictive algorithms in immigration and border control. Predictive algorithms use historical data to make predictions about future migration patterns and trends. However, if the historical data is biased or incomplete, the algorithm may perpetuate pre-existing biases and inequalities, leading to the entrenchment of existing marginalizing patterns in the context of migration. For example, if an algorithm is trained on data that disproportionately represents certain groups or regions, it may be more likely to identify those groups or regions as high-risk for migration, even if the actual risk is lower. This can lead to discriminatory policies that unfairly target certain groups or regions, and hinder efforts to address systemic issues related to migration. 
  • Uncertainty: Another challenge is the uncertainty problem, where the accuracy level of anticipations can vary significantly, mainly due to the data that is available and its different conceptualizations, shock events, unpredictable changes to migration drivers, and large variations in how migrants respond to these changes. For instance, the research article Predicting Spanish Emigration and Immigration utilizes a gravity type model to predict bilateral migration flows up to 2100. However, the predictions showcase large variations in projections beyond 2050, emphasizing the challenge posed by the inherent uncertainty in anticipatory methods.
  • Legitimacy of Preemptive Intervention: Additionally, there is the preemptive intervention problem, where predictions about future migration events remain unverifiable until those events actually occur, and decisions or interventions based on those predictions may result illegitimate. In the context of anticipating migration for policy making, this raises important concerns about accountability and transparency in decision-making processes. In addition, techniques such as machine learning can be used to model migration flows, yet the models and variables analyzed therein can be opaque, and the decisions made based on them can be hard to trace back, verify, and account for. 
  • Group Privacy: Finally, using data-based systems to anticipate migratory patterns and design policy interventions presents the challenge of preserving group privacy. While there are an ever growing number of international policies and frameworks addressing individual data protection and privacy, those hardly take into account people’s right to privacy as a group. The idea that groups have a right to privacy in society is based on the notion that certain groups may be easily identified and targeted, despite the relative anonymity of the individuals making up that group. This is especially important to prevent the risk of discriminatory surveillance, highly relevant in the field of migration policy, and especially when employing anticipatory methods within it. An example highlighting the need to raise awareness on the importance of group privacy is the potential use of digital advertising trace data from Facebook’s Marketing API to estimate daily subnational population sizes in Ukraine, potentially disaggregated by age and sex. While the data can be anonymized and aggregated to preserve individual privacy, it is important to highlight that protecting the privacy of each group member will not, in fact, protect the privacy of the group. This emphasizes the pressing necessity for a more comprehensive assessment of responsible data practices, one that includes group privacy, especially when utilizing such data for policy-making decisions that impact entire groups, so as to ultimately ensure the privacy and dignity of all individuals within these populations.
  • Self-fulfilling Prophecies: Algorithmic predictions have the capacity to shape the future they aim to anticipate, leading to a self-fulfilling prophecy dynamic that can ultimately lead to inertia in the system. This feedback loop between predictions and migration behaviors can produce unintended outcomes and indirectly perpetuate unwanted patterns. Given the nascent development and use of these tools, our repository does not include any explicit real-world examples of this challenge. However, it seems worth acknowledging the potential for this issue to arise as they mature.
  • Politicization of Predictions: One significant challenge can arise when predictions about migration are politicized and used for political purposes. Instead of employing these predictions to inform evidence-based policymaking and address migration challenges, political actors may manipulate and exploit them to serve their own agendas. It is thus important to develop auditing and monitoring mechanisms that ensure that the anticipation insights are not politicized. This requires ongoing scrutiny, evaluation, and refinement of the methodologies used in generating predictions. Given the nascent development and use of these tools, our repository does not include any explicit real-world examples of this challenge. However, it seems worth acknowledging the potential for this issue to arise as they mature.
  • False Positives and Negatives: False positives and false negatives are important considerations within the realm of anticipatory methods, particularly in the context of migration policy. False positives occur when predictions are made for events or risks that do not come to pass, potentially leading to unwarranted actions or resource allocation. Conversely, false negatives arise when actual migration events or risks go undetected, resulting in inadequate preparedness or support. Ongoing model refinement, data quality enhancement, and feedback mechanisms are integral to reducing these errors. Furthermore, ethical and legal safeguards, including transparency, accountability, and mechanisms for challenge and redress, are essential for ensuring the responsible and equitable use of algorithmic prediction in the field of migration policy. As for the Self-fulfilling Prophecies and the Politicization of Algorithmic Predictions challenges, our repository does not include any explicit real-world examples of the politicization of algorithmic predictions. However, it seems worth acknowledging the potential for this issue to arise as these methods mature and are adopted more broadly. Doing so, indeed, may allow for proactive and timely strategies to tackle the challenge.

Implementation challenges

  • Interdisciplinary Collaboration: Effectively utilizing anticipatory methods often requires collaboration across multiple disciplines and sectors. Integrating expertise from fields such as data science, social sciences, migration studies, and policy-making is essential but can be challenging due to differences in terminology, methodologies, and approaches. The FES Geneva, the International Organization for Migration (IOM) and Global Future invited four independent scenario teams composed of participants from civil  society, academia, migrants, and various regions were brought together over an 18-month period to explore potential migration scenarios by 2030, resulting in the development of scenarios through workshops, webinars, and interviews with key individuals. Similarly, the EU Joint Research Centre (JRC) delineates four possible future migration scenarios towards 2030, providing a set of interactive tools that stimulate forward-looking and strategic discussions and that can be used to involve various actors that shape migration policymaking and research in a collaborative and interdisciplinary way.
  • Data Availability and Quality: Anticipatory methods heavily rely on data, and their effectiveness is contingent upon the availability and quality of relevant data sources. Obtaining accurate, timely, and comprehensive data can be a challenge, particularly when dealing with complex and dynamic phenomena such as migration. Data gaps, inconsistencies, and biases can limit the accuracy and reliability of anticipatory models and predictions. The US National Intelligence Council Global Trends 2040 report methodology uses their own previous editions and extensive research and consultations to identify future trends. Faced with the lack of data availability, they are generating the data themselves, ensuring its quality, to envision possible futures.
  • Capacity Building: Implementing and adopting anticipatory methods in migration policy necessitates the development of adequate capacity within institutions and among policymakers. This challenge involves providing training and resources to enhance data literacy, analytical skills, and the ability to interpret and utilize anticipatory insights effectively. For instance, the Mixed Migration Centre (MMC) implemented exercises to build capacity for various stakeholders to effectively plan mixed migration including crisis scenarios. Capacity building efforts may also focus on promoting a culture of evidence-based decision-making and fostering a multidisciplinary approach to understanding and addressing migration dynamics, ultimately moving from anticipatory exercises to policy making processes.
  • Stakeholder Engagement: Engaging relevant stakeholders is crucial for the successful integration of anticipatory methods in migration policy. This challenge involves fostering collaboration and communication among policymakers, researchers, civil society organizations, and affected communities. An example of this is Hilando Futuros - Hivos/ILDA , where a group of textile workshops where different stakeholders, including migrants, meeting to reflect on the present and future of migration processes through the creation of an embroidered textile piece based on the narratives and stories of individuals who have chosen to migrate from other countries to Uruguay. Stakeholder engagement facilitates the sharing of knowledge, data, and expertise, ultimately (a) ensuring that anticipatory methods are informed by diverse perspectives and contextual insights and (b) potentially enabling transparency and accountability in migration policy decision-making processes.

Next week we will present the two primary categories within our proposed taxonomy of innovative anticipatory methods: Emerging Forecast Methods and Emerging Foresight Methods. Stay tuned as we continue to unravel the world of anticipation in migration policy!

We would like to express our gratitude to members of the BD4M Alliance Martina Belmonte (JRC), Damien Jusselme (IOM), and Alina Menocal Peters (IOM) for their valuable reviews of this piece before its publication.

The cover image of this blog was created using DALL-E.

Back to the Blog