Disaster risk is usually understood as the interaction of three things: hazard, exposure and vulnerability. Hazard captures the likelihood and severity of damaging events; exposure captures the people and assets in harm’s way; vulnerability captures how susceptible those people, assets and systems are to damage. A region facing severe hazards should register greater risk, as should one with more people or assets exposed to them.
Bringing these dimensions together in a single index is intended to provide a more complete measure of risk. Yet their presence in the same formula does not necessarily mean that each has a comparable influence on the resulting score. That influence also depends on how much each component varies across states. If one varies enormously while the others vary only modestly, it can drive much of the difference in overall risk scores even without being assigned greater importance in the formula.
A formally multidimensional index can therefore produce results dominated by one of its components. When those results feed into the distribution of public funds, the statistical behaviour of the underlying variables becomes part of the allocation mechanism itself.
The Hidden Weight of Population
Hazard measures are often constructed by aggregating several normalised hazards, which can compress differences between states. Vulnerability measures based on aggregated socio-economic indicators can similarly occupy a relatively narrow range. Exposure, represented primarily by population, varies enormously.
The consequence is that population can acquire an implicit weight beyond what is apparent from the index’s formulation. Its greater variability gives it greater influence over differences in the resulting scores.
Population is a legitimate measure of exposure, and a densely populated state facing disaster risk obviously requires substantial resources. The concern is whether hazard and vulnerability retain enough measurable variation to materially influence the assessment alongside it.
When Disaster Risk Shapes Public Spending
The Sixteenth Finance Commission framework brings this statistical characteristic directly into public finance. Its Disaster Risk Index (DRI) is incorporated into allocations for both the State Disaster Response Fund (SDRF) and the State Disaster Mitigation Fund (SDMF). For 2026–31, the recommended allocations amount to ₹1,63,521 crore for SDRF and ₹40,880 crore for SDMF, or ₹2,04,401 crore in total.
Analysis of the framework reveals a striking pattern: exposure, represented primarily by population, explains nearly 65 per cent of the cross-state variation in observed SDRF allocations. The SDMF displays a similar population-oriented pattern.
The allocation framework also incorporates a baseline-spending component based on states’ previous expenditure, so the 65 per cent figure does not imply that population mechanically determines the same share of allocations through the DRI. It does show how closely the eventual distribution tracks exposure. Where exposure varies far more than the other components, the physical variation in disaster risk may have less influence on the resulting distribution than the structure of the index suggests.
The representation of hazard therefore becomes central to how effectively the index differentiates risk across states.
Making Hazard More Visible
The proposed Enhanced Disaster Risk Index (EDRI) tests this by changing how hazard is represented while preserving the basic multiplicative structure of the existing risk index. It parameterises hazard through three physically interpretable components: intensity, frequency and fragility. Intensity captures the severity of an event, frequency captures how often it occurs, and fragility captures physical susceptibility associated with characteristics such as terrain and geology.
The EDRI is a proof-of-principle framework. It shows how greater differentiation within the hazard component changes the resulting geography of relative risk.
Using illustrative weights of 0.4, 0.3 and 0.3 for intensity, frequency and fragility respectively, the relative risk shares of Nagaland, Sikkim, Tripura and Mizoram rise by more than 200 per cent, while Uttarakhand’s rises by 81 per cent. At the other end, Rajasthan’s relative share falls by 36.5 per cent and Madhya Pradesh’s by 28.6 per cent; Uttar Pradesh and Maharashtra fall by 13.4 per cent and 12.3 per cent respectively.
These changes illustrate how strongly measured risk can respond to what the hazard component is capable of distinguishing. Himalayan and north-eastern states facing combinations of seismicity, landslides, extreme rainfall and fragile terrain become more prominent, while coastal states receive greater recognition of cyclone and storm-surge risks.
The allocation formula itself has not changed. What has changed is what the index is capable of seeing.
Index Design Is Policy Design
The exercise points to a wider question in the use of composite indices. Policymakers usually scrutinise which variables an index contains and what weights they have been assigned. The variability of the underlying data can create another, implicit hierarchy of influence that is less immediately visible.
Any index used to allocate public resources should therefore be tested for which variables actually drive its results. A variability audit could examine the dispersion of individual components and test how strongly each influences rankings or allocation shares. Such scrutiny would make visible the effective weights that emerge from the data.
For disaster risk, the same principle supports greater use of physically measurable and traceable hazard indicators, drawing on data from institutions such as the India Meteorological Department, Central Water Commission, Geological Survey of India and National Remote Sensing Centre. The aim is to preserve meaningful differences in hazard intensity, recurrence and physical susceptibility across states.
Vulnerability requires similar scrutiny. While the EDRI demonstrates the effect of greater differentiation in hazard, vulnerability remains relatively compressed in the existing framework. Infrastructure resilience, adaptive capacity, demographic characteristics, access to resources and socio-economic inequalities could provide a more differentiated picture of susceptibility.
The larger lesson extends beyond disaster finance. Composite indicators increasingly influence decisions about development, resilience, environmental risk and public investment. An index may contain all the relevant dimensions and assign them apparently reasonable weights, while the statistical characteristics of its underlying data produce a very different hierarchy of influence.
Measurement choices can therefore shape policy priorities long before an allocation formula is applied.
Before asking whether an index has the right weights, policymakers should ask a more basic question: What weights have the data quietly given it?