The seawall ends at a survey marker about four hundred metres north of the temple, because that is where the ward ends. North of the marker the beach has lost, by the measurements the fisheries department itself publishes, between eleven and eighteen metres of width since the wall was completed.
South of the marker, behind the wall, the houses are fine.
This is not a surprise to anybody and it is not a failure of engineering. It is what a seawall does. Waves that would have spent themselves across a sloping beach hit a vertical face, reflect, and scour — at the toe of the wall and, more consequentially, at the point where the wall stops and the unarmoured shore begins. Every coastal engineer in Kerala knows this. It is in the textbooks. It was in the environmental assessment.
The wall was built anyway, and the reason is not corruption and not ignorance. It is that the unit of decision is a ward.
Follow the decision. A ward member is answerable to the households in her ward, whose houses are being taken by the sea, in some cases at a rate of a metre or two a year, which is not an abstraction — it is a wall of somebody's house standing in water at high tide. She can obtain funds for coastal protection. The funds are administered per project, and a project is defined by a stretch. She builds the stretch. Her households are safe. The erosion moves north into a different ward, whose member will, in due course and with equal legitimacy, apply for her own wall.
Nobody in that sequence has done anything indefensible. The alternative available to her was to decline to protect her own households on the grounds that the damage would move somewhere else, which is not a thing any elected person has ever done anywhere.
So the coast gets armoured in segments, each segment accelerating the loss at its northern end, and the total length of wall grows every year, and the beach disappears in a pattern that looks like a zip being pulled.
What the beach was is the part that gets left out of the engineering discussion, and it is the part I care about, because it is a market.
A landing beach is where a small boat comes ashore without a harbour. It is where the catch is carried up, where the baskets are set down, where the auction is called, where nets are spread to dry and mended, and where the ice comes on a cycle. It requires width and it requires slope. A wall gives you neither: the water comes to the face of it, and there is nowhere to put a basket down.
Take the beach and the small boats do not stop fishing. They move to the harbour, ten or eighteen kilometres along the coast, and at the harbour the buying is not a beach auction called in fifteen seconds among people who know each other. It is a commission agent with a shed and a weighing arrangement and a credit line, and a fisherman landing there is landing into somebody else's market as a stranger, with fuel already spent getting there.
That is a transfer of margin from the boat to the agent, caused by a wall, and it appears in no assessment of the wall because the wall was assessed for erosion. The fisheries department publishes the erosion figures. Nobody publishes what happened to the price at the two landing centres that closed, which I have asked about twice, and which somebody at the district could work out in an afternoon from the auction records the cooperative already keeps.
The households whose houses the wall protects and the households whose landing it removed are, in a fair number of cases, the same households. That is not a paradox and it is not an argument against the wall. It is what happens when a decision is made against one variable in a place where everything is coupled.
The technical answer exists and it is not new. Sediment moves along this coast in a direction, in quantities that are measurable and have been measured, and the correct scale of intervention is the cell that sediment moves within, which is tens of kilometres and crosses several wards and in places a district boundary. At that scale you can do things a wall cannot: nourish, manage the offshore, decide which stretches to armour and which to let breathe, and — the part nobody says out loud — decide which houses will not be defended.
That last item is why the cell-scale plan does not happen, and I want to be honest that it is not merely an administrative failure. A ward-level wall gives a specific family a specific wall. A sediment-cell plan gives some families a wall and tells others, in writing, that theirs is not coming. No official wants their signature on the second document, and I am not sure I would want mine on it either.
But the current arrangement makes the same decision. It simply makes it by sequence rather than by choice: the ward that builds first is defended, the ward that builds second is defended at the first ward's expense, and the ward at the end of the line — which is always the poorest one, because a wall requires a member who can get funds — is defended by nobody and told nothing.
There is a stake in the sand about eighty metres north of the marker, put there by a man whose house is now the closest one to the water. He drove it in himself and he checks it. He told me it had been fourteen metres from the house in 2021.
The unit of decision being a ward while the sediment cell is tens of kilometres — that is the same mismatch as a watershed managed by district. Every coastal and river system has this and no administrative geography has ever matched a physical one.