A bad summer is not an annual cost. It is a ten-year charge.

The single-year framing is what makes coastal water look unaffordable to fix and cheap to ignore. Spread the same numbers over ten years and the sign of the answer changes.

Year one: the visible bill

Year one is the only part of this most authorities have data on, because it is the part that runs through the general ledger. Emergency water-quality sampling at a frequency nobody budgeted for. Signage, barriers, extra beach cleansing, additional refuse rounds for what washes ashore and rots. Overtime across environmental health. Communications, then legal, then compensation correspondence.

It is uncomfortable but tractable. The finance team can put a figure on it by October, and usually does.

What year one also produces is a number that gets quoted for the rest of the decade, and it is almost always too small. NOAA Fisheries recorded $7.2 million of direct harvester losses from Maine's 2005 shellfish closures. The total income loss to the region from the same event was $49.46 million. The direct figure — the one an authority can count — is roughly a seventh of the real one.

Years two to four: the season stops booking

Tourism does not respond to water quality. It responds to the memory of water quality, which lags by two seasons and recovers more slowly than it falls.

The scale is not theoretical. A University of Central Florida study assessed the 2018 Florida red tide season at $2.7 billion across the state, split roughly $1.27 billion in the south-west and $1.35 billion in the south-east, and its authors said plainly that in revenue terms red tide was worse for coastal tourism businesses than hurricanes. We have that study through news reporting rather than the journal itself, so we mark it indicative rather than established — but even discounted heavily it dwarfs anything in the year-one column.

What happens inside those years is mundane and cumulative. An operator drops the destination from a brochure. A conference bureau loses a rebid it never announces. A hotel discounts to hold occupancy, then finds the discount is now the price. None of it appears as a water cost. All of it is one.

Years five to ten: the structural losses

By year five the problem has stopped being about individual seasons and started rearranging the local economy.

The fleet contracts. Boats are not sold in a dramatic single decision; a skipper simply does not replace an engine, and a berth stays empty. NOAA's coastal science programme puts Gulf commercial fishing at around a billion dollars a year and lists hypoxia among the conditions eroding the habitat behind it. Habitat loss is slow, and so is the corresponding loss of a landing chain, a processor, an ice plant and the jobs attached to all three.

Property adjusts. NOAA Fisheries attributes $10.05 million in lost shoreline property-value services to the single 2014 Lake Erie event. Habitat loss compounds it: the 2011 superbloom in Florida's Indian River Lagoon carried an estimated $235–470 million potential loss through seagrass damage alone, and seagrass is nursery ground, sediment stabiliser and water-clarity engine at the same time.

And the labour market moves. Ocean-economy jobs are unusually place-bound — the OECD's 2010 baseline put direct global employment at about 31 million, with industrial capture fisheries over a third of it and maritime and coastal tourism almost a quarter. Those jobs cannot relocate inland with the household. When they go, the household goes.

What the ten-year ledger keeps missing

1. The recurrence, not the event A single event is a shock. The same fortnight failing in six years out of ten is an operating condition, and it should be modelled as one — with a frequency, not an anecdote.
2. The indirect multiplier Maine's 6.9× is not a universal constant, but the direction is universal. Anyone budgeting from direct losses alone is planning against a fraction of the exposure.
3. The asymmetry of reputation A coastline loses its reputation in one bad August and regains it over four clean ones. The recovery is the expensive half, and it is not on anyone's cost centre.
4. The counterfactual nobody writes down Doing nothing is a spending decision. It is simply one that never requires a paper, a vote or a signature.

Why the upstream answer is a fifteen-year answer

None of this is an argument against fixing the catchment. Nutrient reduction on land is the only permanent solution to nutrient-driven blooms, and every serious coastal authority should be pursuing it.

It is an argument about time. NOAA-supported scientists measured the 2025 Gulf of Mexico hypoxic zone at about 4,402 square miles — the fifteenth smallest in thirty-nine years of measurement, and still enormous. The five-year average is 4,755 square miles against the interagency task force's target of under 1,900 by 2035. That is decades of coordinated federal, state and agricultural effort, and the gap is still more than double.

In the Chesapeake, one of the most intensively studied and managed estuaries on earth, the 2025 dead zone still reached a daily maximum hypoxic volume of 10.1 cubic kilometres — close to the long-term median — and totalled 462 cubic-kilometre-days across a season that ran to sixty sustained days of hypoxia. That was a normal year. Normal is the problem.

Those are not failures of policy. They are the honest timescale of catchment work. The question for a coastal authority is what happens to the fisheries, the beaches, the intakes and the season during the fifteen years that work takes — and whether those years are simply written off.

How to build the number for your own coastline

Nobody outside your authority can do this arithmetic for you, and any supplier who offers to has already told you something about their evidence standards. What we can do is set out the shape of it.

Take the last ten years. Mark the months in each year when the water failed and how it failed. Attach the direct costs you can evidence, then apply a multiplier drawn from a documented comparable rather than from optimism. Add the season revenue that did not arrive in the two years after each event. Add the difference between what your shoreline property base did and what the regional average did. Then look at the total, divide by ten, and compare that annual figure with the cost of holding treatment capacity in the water for the eight weeks a year that actually matter.

Some coastlines will come out of that comparison and decide to do nothing. That is a legitimate answer, arrived at properly, and it is a far better position than never having run the numbers.