Floods as Social and Economic History

Floods in Mithila, Vajji and Anga are not merely episodes of excess water. They are recurring social and economic processes in which rainfall, river morphology, embankments, drainage, settlement location, landholding, caste, gender, transport, public administration and household assets interact. The same depth of inundation can fertilise one field, destroy another crop, isolate a school, erode a homestead, interrupt a market road or push a land-poor household into debt. A socio-economic history must therefore ask not only where water spread but who was exposed, what was lost, who could recover, and how repeated floods altered livelihoods and institutions over time. North Bihar’s flood history is also plural. Himalayan tributaries can rise early; Burhi Gandak and other systems may peak later; the Ganga can become a high master drain and slow the evacuation of local water; embankments may breach; drainage congestion may prolong inundation; and the Ganga corridor in Anga can experience high stages, erosion and backwater effects even when the northern tributaries behave differently. This chapter therefore treats major post-independence flood years as historically distinct events, while also examining the quieter annual accumulation of crop loss, debt, migration, interrupted schooling, public expenditure and adaptive knowledge. 78.1 Floods are produced by water, exposure and vulnerability together Hydrology determines where and how water moves, but disaster losses are socially distributed. A pucca house on raised ground, a thatched dwelling beside an embankment, a tenant’s standing paddy crop and a trader’s stock in a market town do not face identical risks. Land title, savings, livestock ownership, caste location, access to boats, roads, credit, health care and public relief shape what happens after inundation begins. This distinction helps explain why flood mortality, displacement and indebtedness can remain concentrated even when engineering works reduce average water levels in some protected tracts. Flood history must therefore combine river records with evidence on households, labour, markets and institutions. 787787 GAJENDRA THAKUR Figure 308 — How a flood becomes a social and economic disaster 78.2 Bihar’s flood geography is exceptionally extensive but measurement depends on method Government descriptions have long classified about 68,800 square kilometres, roughly 73 per cent of Bihar’s geographical area, as historically flood affected or flood prone. A newer Central Water Commission satellite-based assessment for 1986–2022 identifies about 2.914 million hectares actually affected by observed inundation over its study framework. These are not contradictory numbers: one is a broad historical hazard category, while the other is derived from a particular remote-sensing method and period. The difference is itself historically important because changing measurement technologies alter how the state sees flood exposure, prioritises districts and allocates management resources. 78.3 Himalayan catchments make north Bihar a transboundary floodplain Most major north-Bihar rivers receive water and sediment from catchments extending into Nepal and the Himalaya. Heavy rainfall far north of the political boundary can therefore raise water levels in Mithila and Vajji even when local rainfall is moderate. The flat alluvial plain then slows drainage and encourages channel migration, overbank spreading and waterlogging. Flood governance is consequently transboundary in hydrological fact even where relief and compensation remain organised by Indian administrative boundaries. Forecasting rainfall and river stages in Nepal, maintaining barrages and embankments, and transmitting warnings downstream have become parts of everyday economic security for villages, market centres and transport corridors. 78.4 Flood timing changes through the monsoon season Bihar’s Flood Management Information System notes a recurring seasonal sequence: Bagmati, Kosi and Kamla systems can produce early floods in May and June; Burhi Gandak commonly becomes more important by mid-July; and by September a high Ganga can make drainage more difficult by raising the receiving water level of the master river. This sequence matters economically. Early inundation may affect transplanting and seedbeds, mid-season floods can destroy standing paddy or isolate labour markets, and late water can delay harvesting and rabi preparation. Households may therefore experience several distinct flood shocks in one monsoon without a single record-breaking discharge. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II Figure 310 — Typical seasonal sequence in north Bihar 78.5 Inundation, breach, erosion and waterlogging are different hazards The word ‘flood’ can conceal several mechanisms. Overbank inundation may rise and recede gradually. An embankment breach can send deep, fast water into settlements accustomed to apparent protection. Riverbank erosion can remove land permanently rather than temporarily wet it. Drainage congestion can keep low-lying fields submerged after river levels begin to fall. Flash floods from smaller Himalayan catchments create another pattern. The economic consequences differ: erosion threatens title and settlement continuity; waterlogging reduces cropping intensity; a sudden breach produces acute evacuation needs; and predictable seasonal spreading may sometimes be incorporated into cropping, fishing or grazing strategies. 78.6 Sediment can be both an agricultural resource and a destructive burden Alluvial rivers carry silt that historically renewed soils, raised fields and supported riverine agriculture. Yet sediment is not automatically beneficial. Coarse sand deposited after a breach can bury fertile topsoil and render fields difficult to cultivate for years, while fine silt may be welcomed. Embankments can alter where sediment settles, raising riverbeds or trapping water on the protected side. The economic meaning of sediment therefore depends on grain size, depth, timing and land use. Local debates over embankment openings or cuts cannot be understood without recognising this double character: the same river material may be valued as fertility in one place and experienced as land loss in another. 78.7 Post-independence flood policy inherited an engineering-centred vision The early decades after independence treated barrages, embankments and canals as central instruments for controlling rivers and promoting development. Chapter 77 examined this infrastructure directly. Social history adds another layer: the expansion of protected areas encouraged denser settlement, road building and intensified cultivation behind embankments, while maintenance failures or breaches could then produce especially severe losses. This does not mean that engineering created every flood problem. Rather, infrastructure changed exposure. Once households, schools, markets and crops came to rely on structures, the consequences of overtopping, drainage failure or breach became part of the region’s political economy. 789789 GAJENDRA THAKUR 78.8 The 1987 flood became a benchmark in public memory and official statistics The 1987 flood remains one of the largest post-independence benchmarks for Bihar. The Flood Hazard Atlas retrospective series records about 4.2 million hectares and 24.5 million people affected in that year. Later flood reports repeatedly use 1987 high-flood levels as comparison points at gauges such as Jhanjharpur and Bhagalpur. Its importance therefore lies not only in immediate losses but in the way it became a reference year for engineering design, administrative memory and household recollection. When later gauges approached or surpassed 1987 levels, the comparison communicated risk to officials and residents more effectively than an abstract discharge number. 78.9 The long flood season of 1998 showed the cost of repeated pressure In 1998 rainfall and high discharges placed prolonged pressure on embankments across Burhi Gandak, Bagmati, Adhwara and Kosi systems. Bihar’s flood-history records report 381 deaths, substantial crop damage and repeated emergency flood-fighting works. The historical significance is the duration of exposure. A long season multiplies costs even without one spectacular breach: labour is diverted to protection works, roads remain unreliable, fodder deteriorates, diseases spread, school attendance falls and farmers lose opportunities to replant. Repeated smaller shocks can therefore rival a single catastrophic event in cumulative household impact. 78.10 The 2004 flood was a tributary-basin catastrophe even with a relatively low Ganga The 2004 flood demonstrates that a low master river does not guarantee safety. Heavy rainfall in Himalayan catchments drove Bagmati, Burhi Gandak, Kamla-Balan, Bhutahi-Balan and the Adhwara group to exceptional levels; the official flood history notes that Bagmati at Dubbadhar exceeded its previous high by about 1.18 metres. Vast parts of north Bihar were inundated through multiple embankment breaches. FMISC records roughly 23,490 square kilometres badly affected and 885 deaths in the broader official loss series. The event shows why Mithila’s flood history cannot be reduced to the Kosi alone. 78.11 The 2007 flood revealed the scale of a multi-district relief emergency The 2007 monsoon again produced widespread flooding in north Bihar. The National Institute of Disaster Management’s case study estimated about 19 million people affected across 20 districts and roughly 450 deaths, with damage on the order of ten billion rupees. The response mobilised evacuation, camps, food distribution and aerial relief, but post-event assessment also identified weaknesses in standard operating procedures, sanitation, rehabilitation and social security. Flood relief thus became a test of administrative capacity beyond engineering departments. The episode helped strengthen the later shift toward disaster- management planning, dedicated response forces and pre-positioned shelters and boats. 78.12 The 2008 Kosi breach was a different kind of disaster On 18 August 2008 the eastern Kosi afflux embankment failed near Kusaha in Nepal, allowing the river to move into an older channel belt and inundate large parts of Supaul, Madhepura, Saharsa, Araria and Purnia. World Bank recovery documents describe about 3.3 million people affected, roughly one million evacuated, around 460,000 people temporarily sheltered in 360 relief camps, and more than 500 deaths. Unlike ordinary overbank flooding, this event combined embankment failure, channel avulsion, deep sediment deposition and destruction of roads and bridges. Its recovery therefore required rebuilding settlements and connectivity, not merely distributing short-term relief. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II Figure 309 — Four flood events, four different historical patterns 78.13 The Ganga makes Anga’s flood history distinct from the northern tributaries Anga’s Bhagalpur–Munger corridor experiences the Ganga as both trunk river and receiving channel. Bihar flood records note that Bhagalpur’s Ganga gauge exceeded the 1978 high-water record in 2003, while the 2011 season again brought exceptionally high Ganga levels after decades of comparison with the 1987 benchmark. Here flood history includes riverbank erosion, diara settlement, sand deposition, ferry and bridge disruption, and the interaction of Ganga stages with tributary drainage. Bhagalpur’s repeated appearance in district flood-frequency tables also cautions against treating flood vulnerability as exclusively a north-of- Ganga phenomenon. Table 78.1 — Selected post-independence flood events and what they reveal Event Defensible evidence used here Historical significance 1987 Flood Hazard Atlas retrospective Benchmark year for later gauge series: about 4.2 million ha and 24.5 comparisons and public memory. million people affected. 1998 FMISC records 381 deaths plus major Shows cumulative cost of repeated crop and public-property losses pressure without one single iconic during a prolonged season. breach. 2004 About 23,490 sq km badly affected; Tributary-basin catastrophe even extreme Bagmati/Kamla/Adhwara though the Kosi remained broadly levels; FMISC loss series reports 885 normal. deaths. 2007 NIDM estimated about 19 million Large multi-district relief emergency people affected in 20 districts and that exposed institutional gaps. roughly 450 deaths. 2008 Kosi World Bank: about 3.3 million Embankment breach and channel affected; around 1 million evacuated; shift required reconstruction of more than 500 deaths. settlements and connectivity. 78.14 Floods reorder the agricultural calendar Agricultural loss depends as much on timing as on depth. Floodwater arriving before transplanting can delay the kharif season but leave time for re-sowing. Water arriving after significant expenditure on seed, 791791 GAJENDRA THAKUR fertiliser and labour can destroy sunk investment. Prolonged inundation may prevent a second crop by delaying land preparation into the rabi season. Farmers respond through crop choice, staggered sowing, short-duration varieties, raised seed storage, livestock mobility and diversification into wage work. These practices are not evidence that floods are harmless; they are adaptations developed because flood risk is recurrent and household survival cannot wait for complete hydraulic control. 78.15 Crop loss rapidly becomes a credit problem A destroyed crop removes both expected food and expected cash. Small and marginal cultivators may then borrow for consumption, seed, livestock replacement and house repair at the same time. Formal crop credit can be difficult to restructure quickly where land records are disputed or tenancy is informal. Moneylenders, traders, relatives and self-help groups therefore become part of the post-flood financial landscape. Repeated floods can convert a temporary shock into chronic debt, especially when the next planting season begins before compensation arrives. The economic history of flooding is consequently also a history of interest, collateral, delayed repayment and household decisions over which assets can be sold. 78.16 Livestock losses affect food, traction, savings and women’s work Cattle, buffaloes, goats and poultry are not simply agricultural inputs. They are mobile savings, sources of milk and manure, and in many households central to women’s daily work and income. Floods create fodder shortages, animal disease, drowning risk and the need to evacuate animals separately from people. Relief planning increasingly recognises livestock shelters and fodder, but older accounts often recorded human casualties more systematically than animal losses. When a household loses livestock, recovery may take longer than rebuilding a wall because breeding stock and milk income cannot be instantly replaced. This is especially significant for land-poor households whose principal productive asset may be an animal rather than acreage. 78.17 Fisheries and wetlands complicate the language of pure loss Seasonal water also creates economic opportunities. Floodplain wetlands, chaurs, ponds and temporary connections between rivers and lowlands support fisheries, aquatic plants and grazing. Fishers and boatmen may gain work even while cultivators lose crops. Yet benefits are unequally distributed because access to ponds, capture fisheries and landing sites may be controlled by leases, caste occupations or local power. Pollution, prolonged stagnant water and sudden embankment drainage can also damage fisheries. A balanced history therefore recognises flood-dependent livelihoods without romanticising inundation: the same flood can expand fish habitat while destroying houses and food stocks. 78.18 Riverbank erosion turns a temporary hazard into a land-rights crisis Inundation normally leaves a parcel identifiable when water recedes; erosion may remove the parcel itself. Along the Ganga, Kosi and other shifting rivers, households can lose homesteads and fields to bank retreat and later confront uncertain claims when land re-emerges as chars or diaras. Survey maps, revenue records and physical channels may no longer coincide. Compensation regimes are generally designed more easily for damaged crops or houses than for continuously shifting land. Erosion therefore creates a special form of displacement in which citizenship, residence and land rights remain stable on paper while the physical geography that supported them disappears. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 78.19 Settlement form reflects accumulated flood knowledge Villages across the region historically adapted through raised house plinths, clustered settlement on natural levees or mounds, elevated grain storage, boats, bamboo platforms and knowledge of evacuation routes. Modern roads and embankments can themselves become temporary high ground. Such practices are often labelled ‘traditional resilience,’ but they are costly and unevenly available. A household with resources can raise a pucca plinth or rebuild in brick; a landless family may occupy the lowest available ground. Settlement history thus records both environmental knowledge and inequality in the capacity to act on that knowledge. 78.20 Roads, bridges and markets transmit flood shocks beyond the inundated village A washed-out culvert can isolate villages whose houses remain dry. Milk, vegetables and fish lose value rapidly when transport fails, while food, kerosene and medicine become more expensive in cut-off markets. Traders shift routes, ferries regain importance and wage workers lose days of employment. The 2008 Kosi recovery therefore treated connectivity as livelihood reconstruction: later World Bank completion evidence reported restored roads and bridges benefiting about 1.3 million people and helping around 110,000 farmers regain market access. Infrastructure recovery is economic recovery because rural production depends on movement as much as on fields. 78.21 Schools and health facilities become both casualties and emergency infrastructure Floodwater interrupts schooling through damaged buildings, inaccessible roads and the use of schools as shelters. Children may lose books, uniforms and examination time even when the building survives. Health systems face contaminated drinking water, diarrhoeal disease, snakebite, skin infections, maternal-care disruptions and difficulties moving patients. Relief camps concentrate populations in ways that make sanitation and privacy central concerns. The social cost of a flood therefore continues after water falls: missed schooling and untreated illness can reduce household income and human capital long after crop-loss statistics have been finalised. 78.22 Caste and class shape exposure before the river rises Socially marginalised households are often more likely to occupy erosion-prone banks, embankment interiors, low-lying commons or insecure tenancy because safer land carries higher prices and stronger claims. Research on Bihar’s flood vulnerability consistently finds that social disadvantage and physical exposure overlap. This does not mean every Dalit or landless household occupies the same hazard zone, but it does mean flood losses can reproduce existing hierarchy. Limited savings, informal work, weaker documentation and dependence on local patrons can slow access to compensation and reconstruction. Environmental justice is therefore not an optional perspective on flooding; it is part of explaining why similar hazards yield unequal recovery. 78.23 Gender changes the meaning of evacuation and recovery Male out-migration is common in many flood-prone districts, leaving women, children and older people to manage livestock, food stocks, care work and evacuation during the monsoon. Relief camps may create problems of privacy, sanitation and safety that are not visible in household counts. Women may lose poultry, stored grain, kitchen gardens and home-based work that damage assessments undervalue because they are not 793793 GAJENDRA THAKUR formally recorded as wage income. At the same time women’s self-help groups and neighbourhood networks can become crucial channels for savings, information and relief. Flood recovery is therefore partly a reorganisation of unpaid and community labour. 78.24 Migration is both a response to floods and a source of resilience Repeated crop failure and seasonal isolation can increase temporary migration toward Punjab, Delhi, Kolkata, industrial centres and nearer towns. Remittances then finance house repair, seed and consumption during later floods. Migration can therefore reduce immediate income vulnerability while creating new care burdens for those who remain. A catastrophic flood may also generate distress migration by exhausting local employment. The direction of causation is consequently circular: pre-existing migration changes household capacity to survive a flood, and flood losses alter subsequent migration decisions. The long flood history of Mithila and Kosi regions is inseparable from this wider labour geography. Figure 311 — Flood shock and the household economy 78.25 Relief is an economic institution, not only humanitarian assistance Gratuitous relief, food packets, community kitchens, tarpaulins, cash transfers, livestock assistance and public works influence who must sell assets after a flood. Timely relief can prevent distress borrowing and preserve seed or animals for the next season. Delayed or poorly targeted relief can deepen inequality. Bihar’s 2007 and 2017 responses show the growing institutionalisation of camps, boats, response forces and community kitchens. Yet entitlement remains dependent on enumeration, identity documents and access to officials. The social history of relief therefore includes queues, lists, local mediation and the political visibility of affected settlements. 78.26 Recovery requires rebuilding networks as well as houses The 2008 Kosi reconstruction experience helped shift policy from replacing individual assets toward restoring systems. Owner-driven housing reconstruction was combined with roads, bridges, livelihoods, flood-management capacity and emergency response. Later project reporting linked restored connectivity to schools, health centres and market access, demonstrating that a house cannot function as a recovered household if roads, irrigation channels or public services remain broken. This systems perspective is especially HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II important in the alluvial plain, where one damaged bridge or drainage channel can affect many villages whose physical dwellings escaped direct inundation. 78.27 Embankments are also sites of local politics and contested risk Flood reports occasionally record deliberate cuts or local interventions in embankments. Such acts cannot be reduced to irrational sabotage. Residents may seek to release trapped water, protect one settlement at the expense of another, or allow sediment-bearing water onto fields. At the same time an unauthorised cut can create severe danger downstream. Embankments therefore redistribute risk and create political questions over whose land is protected, whose drainage is blocked and who has authority to intervene during an emergency. The social history of flood control includes these conflicts over gates, closures, spurs and emergency breaches as well as the formal engineering record. 78.28 Flood information became a new form of public infrastructure From the late 2000s Bihar invested in a dedicated Flood Management Information System and later Flood Management Improvement Support Centre. Satellite inundation mapping, telemetry, river gauges, Nepal hydrometeorological information, rainfall forecasts and web bulletins increasingly allow officials and the public to see flood extent and river status in near real time. This informational infrastructure changes economic decisions: district administrations can pre-position boats, farmers can move livestock, traders can alter routes and households can evacuate earlier. Forecasting does not eliminate loss, but lead time converts information into a protective asset. 78.29 New flood maps require careful comparison with older statistics The historical category of 68,800 square kilometres flood prone remains widely used in Bihar planning, while the Central Water Commission’s 2024 remote-sensing assessment reports about 2.914 million hectares affected by floods in Bihar over 1986–2022 under its satellite methodology. The Flood Hazard Atlas adds spatial frequency by showing which areas were repeatedly inundated across observed years. These products answer different questions. A historian should not treat the smallest modern mapped extent as proof that older estimates were wrong, nor treat the older hazard zone as if every hectare floods annually. Measurement history is part of flood history because definitions guide budgets, design standards and public perceptions of risk. 78.30 The durable policy problem is living with floods without normalising avoidable loss Mithila, Vajji and Anga will remain riverine societies. Climate variability, Himalayan sediment, dense settlement and the geometry of the Ganga plain make complete elimination of flooding unrealistic. The historical lesson is instead to distinguish beneficial seasonal water from destructive exposure and to reduce preventable loss through drainage, maintained embankments, floodplain-compatible land use, resilient roads and housing, early warning, health preparedness, social protection, secure land records and rapid livelihood recovery. Floods become disasters when institutions and households cannot absorb the shock. The region’s future therefore depends as much on reducing social vulnerability as on controlling river discharge. 795795