Climate change has been recognized as a key driver of migration and one of the root causes of refugee movements. Its effects are shaping new mobility patterns within and across countries, yet the costs and benefits of these movements are still not fully understood.
What do we currently know about climate migration?
Despite growing body of research, the relationship between climate change and migration remains difficult to analyze, often due to uncertainty regarding not just the magnitude, but even the direction of this relationship. This is partly attributable to the intrinsically complex, highly context-dependent nature of the phenomenon which is characterized by the interplay of different transmission channels and mediating factors. But a lot of the uncertainty also has to do with the existence of severe microdata gaps in origin areas within developing contexts.
The role for household surveys in filling data gaps
Increasing availability of big data offers opportunities for increasing the temporal and spatial granularity of migration data and reducing the costs of tracking hard-to-reach populations. However, this type of data remains vastly untested and sporadically adopted in official statistics and presents yet unresolved ethical and privacy challenges. Moreover, big data alone cannot be the solution to answer multifaceted policy research questions implied by the climate-migration-development nexus (e.g., the ultimate effects of climate and climate-induced migration on household and individual welfare). Given these challenges, the role of traditional migration data sources, especially surveys, should not be dismissed, but amended with the goal to maximize their added value.
In a new working paper, we investigate if and how fit-for-purpose household surveys can help fill the existing data gaps and help better analyze the nexus of climate change and migration.
As a case study we used the World Bank’s Living Standards Measurement Study (LSMS) household survey program with specific focus on the LSMS-ISA program which primarily aims at improving the availability, quality and relevance of agricultural data within multitopic, nationally-representative panel household surveys.
We concluded that, despite some well-known limitations (i.e., a typically small migrant sample size due to the ‘rare event’ nature of the migration decision), longitudinal, multitopic, nationally-representative household surveys remain an essential source of information to study the causal linkage between slow-onset climatic changes and internal migration, climate-induced immobility, the mediating role of household-specific transmission channels such as liquidity constraints and assets, and the relationship between migration and in situ adaptation.
In our opinion, this often-overlooked potential of household surveys can be unleashed and further exploited in three main ways:
- Improve survey tools with easily implementable changes to current questionnaires and the use of mixed modes to enhance data quality and timeliness. This includes harmonizing and aligning screening questions to identify different types of migrants with best practices, integrating specific modules on potential migrants and intention-to-migrate, climate risk perceptions, and migration-as-adaptation, as well as the use of phone-based surveys.
- Enhance integration with non-conventional data sources (e.g., remote sensing, mobile phone traces, social media, and other types of big data) and other traditional data sources (e.g., census data, administrative records, and registries) to further increase the temporal and spatial resolution of survey data.
- Broaden the perspective for some key research questions to make better use of existing household survey data and provide complementary solutions to the ‘rare event’ issue. This implies, for instance, strengthening the focus on climate-induced immobility traps and ‘potential’ migrants to complement the (typically small) actual migrant sample.
Taken together, these efforts could boost the quality and volume of climate migration research and allow for a better design and targeting of both migration and adaptation policies.
Interoperability as the way to go
No single data source can resolve all the microdata gaps that currently hinder a more solid understanding of the climate migration nexus. Multitopic household surveys remain an indispensable (and irreplaceable) source of information for untangling the complex dynamics underlying this relationship and assessing differential impacts (e.g., on women). Still, they have two main limitations: an often-small migrant sample size; and limited ability in tracking international migrants and cross-border mobility.
There is an active debate about how to overcome these limitations. Fresh, massive datasets offer an appealing solution to reduce the existing constraints, provided that they are collected regularly, with high quality standards, and can be properly linked to survey data. Fostering integration of household surveys with non-traditional data sources would allow to maximize the advantages of both worlds: the comprehensiveness, longitudinal nature, and representativeness of survey information with the abundance, high-frequency, timeliness, granularity, and traceability of big data, and shed new, perhaps decisive light on crucial policy research questions such as the indirect role of climate change as a push factor for cross-border mobility via internal mobility and urbanization.
Complementarity and interoperability between old and new data sources is, in our view, the way to go and should inspire the future international research agenda. From this perspective, nascent data collaboratives such as the Development Data Partnership offer concrete promises for addressing key challenges related to data governance and quality and taking the climate migration data landscape to the next level.
Join the Conversation