Formalising an Informal Settlement on an Ephemeral River Floodplain
Ombika municipality has 21,400 residents and is growing at 4.1% a year, with 60% of the population under 25. About 6,500 people live in an informal settlement of 900 plots on the floodplain of an ephemeral river; a flood three years ago put 1,100 households under 1.2 m of water for nine days. Most adults work within walking distance in the town's formal jobs and informal trade, and the settlement's shebeens and market stalls sit along its main track. Council has three proposals: (A) in-situ upgrading - gravel roads, water, sewer and raised concrete plinths - at about N$38,000 per plot; (B) relocation to a serviced site 7 km east on higher ground at about N$54,000 per plot, with a school and clinic planned; (C) a hybrid, upgrading the higher 55% of plots now and relocating the rest. The council has one financial year, N$40 million in capital budget, and a legal opinion that it carries liability for flood deaths on land it has approved.
1. Recommend a proposal, using explicit criteria and the numbers given.
Read the worked answer
Criteria: lives at risk (legal liability and flood depth), cost per household, effect on livelihoods, speed within one budget year, and future growth.
Cost test: the 900 plots at N$38,000 = N$34.2 million for full in-situ upgrading - within N$40 million. Relocation of all 900 = N$48.6 million - over budget. The hybrid: 495 plots at N$38,000 = N$18.8 million, plus 405 at N$54,000 = N$21.9 million, total N$40.7 million - marginally over the capital budget and the most complex to deliver before the next rainy season.
Lives test: option A keeps 1,100 households in a known 1.2 m flood zone; raised plinths reduce but do not remove risk, and liability remains. That counts heavily against A.
Livelihoods test: relocation 7 km east cuts walking access to jobs, schools and the market that the settlement's traders depend on; that is the main reason relocation schemes elsewhere are partly abandoned, with people drifting back to the floodplain.
Recommendation: proceed with the hybrid, but sequence it - relocate the 405 lowest plots first (N$21.9 million, affordable in year one), upgrade the 495 higher plots with plinths and drainage in year two - and negotiate one extra budget cycle for the balance. Attach livelihood measures: a minibus subsidy or a trading site at the new location. Reject pure A, and reject pure B as unaffordable and unenforceable.
2. The council fears that relocated households will sell their plots and move back. What should the scheme include to reduce that, and what data would you collect first?
Read the worked answer
Why people return: the floodplain is not irrational housing - it is a low-cost location next to work, customers and social networks. Return happens when the new site adds travel cost without adding income. So design should attack that, not lecture residents.
Scheme elements:
- Title on conditions: allocate plots with a transfer restriction for a set period and a 'beneficial occupation' clause, so a plot cannot simply be sold to an outsider.
- A trading area and market stalls at the new site, built in phase one, so the informal economy relocates with the households rather than being left behind.
- Transport: a subsidised minibus service for the first two years and a footpath/cycle link to town.
- Services early: water and electricity connections before occupation; a school site serviced in year one even if the building comes later.
- A flood-early-warning and evacuation plan for those who stay, plus no new residential allocations on the floodplain.
- Rental rooms for tenants: 6,500 people on 900 plots means many are lodgers, and they are usually missed by plot-based schemes.
Data to collect first: a household register (not plot count) giving income, employment location, tenure and rent; how many residents are tenants; a flood-level map with depth contours for a 1-in-10 and 1-in-50 year event; existing service capacity at the new site; and willingness-to-move survey by block, so phasing follows consent rather than guesswork. Without the register, the scheme will compensate plot owners and ignore half the population.