Migration scenarios: Informing policy, without predicting outcomes

2020-03-12 07:22:42

For policymakers, non-governmental organisations, and civil society more broadly, having reliable information on migration flows has become increasingly important. Within the framework of the CrossMigration project, the International Organisation for Migration’s Global Migration Data Analysis Centre (GMDAC) in partnership with the Netherlands Interdisciplinary Demographic Institute (NIDI) examined the potential and limitations of using expert opinion on future migration. Two of the lead researchers, Jasper Tjaden and Eduardo Acostamadiedo shared some of their observations and conclusions in an interview.

This interview has been slightly edited for length and clarity.

CrossMigration: Firstly, could you share the idea behind your research? 

Jasper Tjaden: We did not want to develop yet another new set of migration scenarios. We first systematically reviewed the literature that is available and found that there are already many scenarios and many of them are quite similar. What we did instead was to take commonalities from existing scenarios and condense them into four “synthesized” scenarios for international immigration to the EU in 2030. We then ask experts how likely they thought each scenario was, how important each scenario would become what it would mean for migration flows. This would allow us to also make an assessment of how useful these scenarios that have already been developed are, how reliable they are, and how useful for policymakers they can actually be.

Eduardo Acostamadiedo: Yes, so in terms of methodology, we started from already-existing research, we looked at the most robust studies and synthesized the migration drivers that are most impactful and most uncertain to create our scenarios. We set the time frame until 2030, because that’s the timeframe in most of the existing scenarios, and this is often related to the SDGs.

CM: Could you share the main conclusion on the scenarios? 

JT:  Experts think that regular migration to Europe will increase. But we mostly see a continuation of the upward trend over the last decade. We see the largest spike in inflows for highly-skilled migrants regardless of the type of scenario, while irregular and forced migration (non-work related) will remain more or less similar to the average levels of the past 10 years. The peak in irregular and forced migration that was experienced in 2015-2016 will not become the norm, in their view.

CM: So, what would you consider the most novel part of your approach?

JT: To begin with, methodologically the study combines two different approaches: migration scenarios and a Delphi expert survey. We tried to have a sample as diverse as possible in terms of expertise for the survey. The experts that we chose had knowledge of European migration issues. The 180 survey respondents were evenly divided in terms of the length of their experience, and we ensured that a full third of them were practitioners, or had more policy experience.

EA: As for the scenario creation, first we reviewed the current literature and identified the most salient variables. Data pointed towards the economic development gap (economic convergence v. economic divergence) and the degree of international cooperation between the EU and the main sending regions. So, we ended up with four scenarios (see matrix below).

EA: In the second step of the study, we used the Delphi method. Experts were asked to fill out a survey that began with the four-scenario matrix. We asked them to estimate what the level for 2030 would be, but also what the variation would be between the 4 scenarios presented. And we didn’t only ask about the implications for the flows, but also what they thought was the most likely scenario.

Then they filled the survey out again, but they were shown the aggregated results and had an opportunity to change their original response.

CM: And did they?

JT: Our research shows that experts don’t agree much, or on much. To many experts, it is just not clear how improved international cooperation or economic convergence between countries would affect migration flows. In addition, the scenarios that have been developed, so our current way of informing policy on migration are too broad and abstract. Even among the experts, there was an incredible degree of disagreement, and only 10% of the experts changed their minds after seeing the opinion of their peers.

EA: This holds to a lesser extent for irregular and labour related flows, but still, the disagreement persists. This has to do, we believe, with the diversity in migration studies. The background of the experts that participated in our study is different, so they bring different assumptions, knowledge and theoretical paradigms to the table.

CM: would you say that this was the main challenge to your study?

EA: Well, there was also the challenge of reducing the subjectivity of the unit of measure that we used. Each expert had to assign points out of 100 to the likelihood of each scenario, but not all experts share the same definition of what merits a 20 or a 40.

JT: And beyond methodology, there was also practicality. This is probably the largest expert opinion survey with migration experts. It was very difficult to get them to respond, even to two, 20-minute surveys. So we had to design a survey that would be attractive enough and meaningful enough and to put the information in a way that could be understood and reacted to in 20 minutes.

EA: We are very grateful that we got to test the survey approach with a smaller group of experts first. We are very grateful to the experts who went to Brussels to help us along in this process. We also want to thank the almost 180 experts that dedicated their time responding to the survey and made the research possible.

CM: Yet, still, there was no agreement, in spite of all your efforts?

EA: That is true. However, this high uncertainty level and this inability of experts to agree on the future isn’t necessarily a weakness, it can be an asset if we understand how to best use these scenarios, and how to apply them to policy.

JT: Yes, so one of the most important findings we have from this study is that the use of scenarios for policymaking is mostly relevant for long term managerial planning, and less useful for short term operational input.

CM: So what do you hope this new view on the use of research to inform policy will achieve?

JT: What we hope to show is that policymakers, researchers and the public don't take scenarios at face value, but that they act as conversation starters. We hope that the tool is used as a discussion tool rather than a tool to predict or make policy about the future.

Policymakers are careful and deliberate in how they use experts, as they should be. Because our research shows that who you invite to an expert group or a meeting or a study makes a difference. The opinions diverge so much that policymakers should be aware of that. Even if you invite many more experts, there little more convergence or consensus. That’s why when talking about the future, we cannot rely on ‘expert predictions’. All of this research and information is meant to give us a more nuanced understanding of the broader lines, rather than giving specific or punctual advice.

In summary, policymakers can, and should, involve researchers, but more to help them broaden their perspective and avoid-short term mistakes. When it comes to migration, you cannot predict the future, but being aware of the complexity of the issues at hand and the types of variables that we deal with can help avoid rash policy decisions.

To learn more about this research in particular visit the Future Migration Scenarios page on our website. 

Ask us