Resilience Measurement for Development
- Introduced by
- FAO (RIMA); Christophe Béné and colleagues (conceptual framing)
- Year
- 2016
- Type
- Latent-variable framework for measuring development resilience
- Family
- Vulnerability and resilience measurement
What does it do?
Resilience Measurement and Analysis is a family of methods for quantifying the ability of households and communities to withstand, recover from, and adapt to shocks and stresses while maintaining or improving their well-being, especially food security.
Exemplified by the FAO's Resilience Index Measurement and Analysis (RIMA-II) and informed by Béné and colleagues' critical conceptual work, it treats resilience as a latent capacity inferred from observable assets, access to services, and adaptive behaviours, estimated statistically and tracked over time to inform and evaluate resilience-building interventions.
The intuition behind the method
You cannot directly see how resilient a household is — resilience reveals itself only when a shock arrives and some families bounce back while others collapse. The challenge is to estimate that hidden capacity before the shock, from the things you can observe: the assets a household owns, its access to markets and services, its safety nets, and its ability to adapt. Resilience measurement treats resilience like an unmeasured trait behind these visible signs, much as a test infers intelligence from answered questions, and combines the signs into an index that predicts who will absorb a drought or price spike and who will be pushed into hunger.
History
Resilience migrated into development from ecology, where C. S. Holling's 1973 work described the capacity of systems to absorb disturbance, and from disaster-risk reduction. As recurrent crises in the Sahel and Horn of Africa exposed the limits of repeated emergency response, agencies sought to measure and build resilience rather than only deliver relief. The FAO developed the Resilience Index Measurement and Analysis approach, publishing the revised RIMA-II methodology in 2016 with its pillar structure and structural-equation estimation. Christophe Béné and colleagues' 2012 IDS Working Paper 'Resilience: New Utopia or New Tyranny?' provided an influential critical and conceptual framing, and a Resilience Measurement Technical Working Group convened by FSIN worked to harmonise approaches, alongside the growth of subjective-resilience measurement.
When to use it
Use resilience measurement when a programme aims to strengthen households' or communities' capacity to deal with recurrent shocks — drought, conflict, price spikes, displacement — and you need to baseline that capacity, target the least resilient, and evaluate whether interventions improved it. It suits food-security and humanitarian-development contexts with household survey or panel data. It is less appropriate where shocks and outcomes cannot be observed, where panel data are unavailable to validate the index against realised resilience, or where the underlying concept of resilience has not been clearly bounded to specific shocks and well-being outcomes, in which case the measure risks being arbitrary.
Limitations and considerations
- Resilience is a contested, normatively loaded concept; different definitions yield different indices, and Béné and others warn it can mask power and structural causes of poverty.
- Robust measurement and validation require panel data linking pre-shock determinants to post-shock outcomes, which are costly and often unavailable.
- Index estimates depend on the choice of pillars, variables, and statistical model, and the latent-variable approach can obscure how the number was produced.
- Objective indices may diverge from people's own sense of their ability to cope, so subjective and objective measures can give different and even conflicting pictures.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Makes an abstract, latent capacity measurable, allowing the least resilient to be identified and targeted before shocks occur.
- Decomposes resilience into pillars and capacities (absorptive, adaptive, transformative), showing which drivers to strengthen rather than just an overall score.
- When built on panel data, it can be validated against realised outcomes and used to evaluate the impact of resilience-building programmes.
- Provides a standardised, comparable framework (e.g., FAO RIMA) used across countries and agencies, supporting cross-context learning and benchmarking.
Applications
- Establishing a resilience baseline before a food-security or livelihoods programme and targeting the least resilient households.
- Evaluating whether resilience-building interventions raised households' capacity to withstand shocks, using treatment-comparison panel data.
- Early-warning and crisis monitoring by tracking resilience capacity and its pillars in shock-prone regions.
- Comparing the resilience profiles of different livelihood or agro-ecological groups to design differentiated support.
- Informing the design of social-protection and safety-net programmes by identifying which resilience pillars are weakest.
How it works
How it works
Measurement starts from a conceptual model of what resilience is. Contemporary frameworks, following Béné and others, distinguish three capacities: absorptive capacity (the ability to cope with and buffer a shock without major loss of well-being, through savings, reserves, and risk management), adaptive capacity (the ability to make incremental adjustments to livelihoods in anticipation of or response to change), and transformative capacity (the ability to make fundamental, often structural, changes when current systems are untenable). Resilience is defined relative to a well-being outcome — typically food security — and to specified shocks, fixing what the measurement must capture.
Because resilience is not directly observable, the analyst selects observed variables that proxy its determinants. RIMA-II organises these into pillars: Access to Basic Services (schools, health, water, markets), Assets (productive and household, including land and livestock), Social Safety Nets (formal and informal transfers and support), and Adaptive Capacity (education, income diversification, coping options). Variables are drawn from household surveys, ideally panel data that also record exposure to shocks and the outcome of interest, so that the determinants can later be linked to realised resilience.
The pillars are combined into a Resilience Capacity Index using multivariate statistics. RIMA-II uses a two-stage procedure: factor analysis or principal-component analysis reduces the variables within each pillar to a latent score, and a multiple indicators multiple causes (MIMIC) structural-equation model estimates resilience as a latent variable driven by the pillars and validated against an observed resilience proxy such as food-security indicators. This yields both an index value for each household and an understanding of which pillars contribute most, while subjective-resilience approaches instead derive the measure from households' own assessments of their ability to cope.
The index is validated by testing whether it predicts the actual outcome — for example, whether higher pre-shock resilience is associated with smaller losses in food security after a shock — using panel data that observe households before and after. Tracking the index and its pillars over time, and across treatment and comparison groups, allows resilience dynamics to be monitored and the impact of resilience-building programmes to be evaluated, closing the loop from measurement to policy.
Common pitfalls
- Measuring 'resilience in general' without specifying resilience to what shock and resilience of what outcome, producing an index with no clear referent.
- Treating the resilience capacity index as validated when no shock-and-recovery (panel) data have been used to test whether it actually predicts outcomes.
- Reporting a single composite score while hiding the pillar contributions, losing the diagnostic value needed to design interventions.
- Adopting resilience framing uncritically as a depoliticised technical fix, ignoring the structural and political drivers of vulnerability the concept can obscure.
Frequently asked questions
What are absorptive, adaptive, and transformative capacities?
Absorptive capacity is the ability to prepare for, mitigate, and cope with the impact of a shock so that well-being is maintained — for example, drawing on savings, reserves, or insurance to weather a drought. Adaptive capacity is the ability to make proactive and incremental adjustments to livelihoods in response to changing conditions, such as diversifying income sources. Transformative capacity is the ability to make fundamental, structural changes — in institutions, systems, or behaviours — when the existing way of life is no longer viable. Together they describe an escalating set of responses from coping through adjusting to transforming.
How does FAO's RIMA estimate resilience if it cannot be observed directly?
RIMA-II treats resilience as a latent (unobserved) variable inferred from observable pillars: Access to Basic Services, Assets, Social Safety Nets, and Adaptive Capacity. It first reduces the variables within each pillar to a score using factor or principal-component analysis, then uses a multiple indicators multiple causes (MIMIC) structural-equation model in which the pillars cause the latent resilience and the latent resilience is reflected in observed outcomes such as food security. This produces a Resilience Capacity Index for each household and shows how strongly each pillar contributes.
What is subjective resilience measurement?
Subjective resilience measurement asks people to assess their own ability to anticipate, absorb, and recover from shocks, rather than inferring resilience from external indicators of their assets and services. Using survey questions about perceived coping ability, it captures dimensions — confidence, aspirations, self-assessed capacity — that objective indices may miss, and can be quicker and cheaper to collect. It is increasingly used alongside objective approaches such as RIMA, since the two can diverge and together give a fuller picture of resilience.
Related methods
Sources
FAO (2016). RIMA-II: Resilience Index Measurement and Analysis-II. Food and Agriculture Organization of the United Nations, Rome.
Open publication referenceCopied.Béné, C., Wood, R. G., Newsham, A., & Davies, M. (2012). Resilience: New Utopia or New Tyranny? Reflection about the Potentials and Limits of the Concept of Resilience in Relation to Vulnerability Reduction Programmes. IDS Working Paper 405. Institute of Development Studies, Brighton.
Open publication referenceCopied.