Do floods still shock the system, or have they become routine enough to barely register? Each year, rivers overflow, embankments give way, homes collapse, and families lose what little they possess. Relief is announced, compensation lists are drawn up, and life moves on. What remains largely unquestioned is whether this relentless repetition reflects natural inevitability or a deeper policy failure.
The Household-level Flood Loss Assessment conducted by Megh Pyne Abhiyan and supported by Tata Trusts, after the 2024 floods confronts this question directly. By documenting losses across 2,290 households in 7 districts of North Bihar, the assessment reveals that flood damage is not only recurring but also predictable, unequal, and embedded in how flood management is designed.
Official flood statistics aggregate damage at district or state level, reducing complex human experiences to numbers on hectares and houses. The household assessment exposes the limitations of this approach. During the second phase of the 2024 floods, the median household loss stood at ₹2.11 lakh, an amount that can erase years of savings for low income families. These losses are not evenly distributed, nor are they short lived.
The composition of household loss is especially revealing. Land damage alone accounted for ₹55.4 crore, or 43.84 percent of total reported losses, while housing repair and reconstruction costs contributed ₹42.51 crore, or 33.6 percent. Together, land and housing losses made up approximately 80 percent of household damage. Yet land erosion, silt deposition, and declining cultivable value remain largely invisible within compensation frameworks, while housing support rarely matches actual repair costs in flood prone areas.
Nearly 87 percent of surveyed households reported housing damage, underscoring how shelter functions not only as a physical structure but also as the primary store of household wealth. Damage to everyday items such as kitchenware, groceries, furniture, and sanitation facilities may seem minor in monetary terms; however, these losses are widespread and can disrupt daily life for months, particularly for poorer households.
The assessment also reveals a vulnerability paradox that challenges conventional interpretations of damage. Certain households, such as those belonging to the general category or residing in pucca houses, reported higher monetary losses. In contrast, Scheduled Caste and Scheduled Tribe households, as well as those residing in kucha structures, experienced losses that were smaller in absolute terms.This highlights that monetary loss figures alone fail to capture the true depth of hardship. For households with limited financial buffers, even smaller losses can be far more difficult to recover from. In this context, higher reported losses may simply indicate that a household had more assets to lose in the first place Nearly 87 percent of surveyed households lived in kucha houses, severely constraining recovery options.
Flood typology deepens this inequity. Breach induced flooding affected the largest number of households and produced the highest aggregate losses. Flash flooding between embankments, often accompanied by erosion and heavy sedimentation, caused extremely high losses for a smaller number of families. Spatial analysis shows that households located outside embankments and those between embankments together accounted for approximately 58% of the surveyed population. The presence of embankments offered little protection during the 2024 Phase 2 floods and instead redistributed risk, creating new pockets of vulnerability.What is framed as protection for some translates into heightened exposure for others.
Coping strategies reveal how survival is purchased at long term cost. About 91 percent of households reduced food intake during the floods. Around 83.67 percent relied on stored food grains, while 75 percent borrowed food from relatives or neighbours. Displacement affected 82.18 percent of households, and 68.43 percent depended on remittances. These are not temporary adjustments but indicators of chronic stress.
Distress asset erosion was widespread. Approximately 35.07 percent of households mortgaged jewellery, 24.93 percent sold or mortgaged livestock, 9.69 percent mortgaged land, and 4.02 percent sold land outright. Among Scheduled Tribe households, rates of land and livestock mortgage were significantly higher, pointing to deeper erosion of future livelihood security. Female headed households, despite reporting lower absolute losses, displayed greater financial precarity and dependence on informal support networks.
Institutional response remains misaligned with lived realities. Despite formal disaster management protocols, households reported short warning periods, uneven relief delivery, and limited involvement of local self-government institutions. Drinking water systems failed, sanitation collapsed, and evacuation was largely self-organised. Nearly 80 percent of households were unaware of or lacked access to any form of flood insurance, underscoring the absence of effective social protection.
Taken together, the assessment reveals a central policy failure. Flood governance continues to rely on aggregate damage counts and uniform responses, while households experience floods through differentiated vulnerability shaped by location, flood type, and social position.
Flood governance must therefore shift from counting damaged assets to understanding household level vulnerability and cumulative loss. Compensation and recovery planning must move beyond houses and crops to include land, livelihoods, and asset erosion, while flood responses must be tailored to specific flood typologies rather than applied uniformly. Without such a shift, floods in Bihar will continue to produce not only waterlogged fields, but locked futures for millions.
(The writer is Managing Trustee, Megh Pyne Abhiyan, which is a public charitable trust working in the water-stressed regions of eastern India, primarily in Bihar and Jharkhand, and intermittently in West Bengal.)
(With inputs provided by Saanjali Verma and Siddharth Patil.)