The Gulf dead zone five-year average was 4,755 square miles as of 2025 against a target of under 1,900 — and the deadline for that target has been moved once, from 2015 to 2035

What is happening?

Every summer a body of water on the Louisiana and Texas continental shelf loses enough oxygen to stress or kill the fish, shrimp and crabs living in it. The zone measured 4,402 square miles when it was surveyed on 2025-07-20 through 2025-07-25, and the five-year running average, which is the metric the goal is actually written against, stood at 4,755 square miles as of 2025 — a figure NOAA describes as more than two times larger than the 2035 target of under 1,900 square miles. The target dates from 2001. Its deadline has been moved once, from 2015 to 2035.

Three days before this document was checked, on 2026-08-05, the scientists who map the zone announced a 2026 measurement of 1,332 square miles, the second smallest since mapping began in 1985. In the same announcement they attribute it to physics rather than to policy: below-average river discharge, and Tropical Storm Bertha, which formed on 2026-07-19 and took an unusual westward path along the northern Gulf coast, making two landfalls in Louisiana. They add that the size represents physical mixing of the water column and is one year of many since the Action Plan was adopted in 2001. Because the goal is written as a five-year running average, a single storm year drags the headline figure toward the target while nothing on any field has changed.

On 2026-06-11 EPA announced that the interim nutrient target set in 2015 had been met for nitrogen, reporting total nitrogen down 28 percent against a baseline period of 1980 to 1996, and missed for phosphorus, reporting total phosphorus up 13 percent against the same baseline. Eight weeks later the announcement from LSU and LUMCON stated that no significant reduction in nitrate loading from the Mississippi River to the Gulf has occurred since 2001. The two statements measure different things over different reference periods, and neither document mentions the other.

Whose problem is this?

Twelve states hold the levers and one state holds the harm. The federal money already spent is large, and it was spent without any state being obliged to reach a number.

Where does this problem end?

The boundary matters here because the goal already exists and the science behind it already exists. What is missing is not a target and not a plan, but anything that happens to anyone when the target is missed.

What is the state now, and what should it be?

Now

The target state is written down and dated. The coastal goal is a five-year running average of the hypoxic zone under 5,000 square kilometers, which is under 1,900 square miles, by 2035, with a 20 percent reduction in nitrogen and phosphorus loading as an interim target for 2025. Set against the 2007 science assessment, the interim target is less than half of what that assessment said the coastal goal needs, and the announced 28 percent for nitrogen is still well short of 45.

How big is it?

The harm channel is stated plainly by the mapping scientists: low oxygen can stress organisms and may cause their death, threatening living resources including fish, shrimp and crabs caught there. The only occupational group in that water that a federal source counts head by head is the Gulf shrimp fleet, at 2,791 crew positions on active federally permitted vessels in 2023, down from 3,426 in 2021 and 3,236 in 2022. NOAA states that federally permitted vessels harvest approximately two thirds of Gulf shrimp landings, so scaling that count gives roughly 4,200 positions across the federal and state water fleet combined.

That arithmetic is not published as an affected population, and it should not be read as one. The federal economic account of this fishery is NMFS-SEFSC-795, issued in March 2026, and a full text search of it returns zero occurrences of hypoxia, dead zone or oxygen. It records revenue falling from USD 489 million in 2021 to USD 221 million in 2023 and more than 1,200 positions lost, and attributes the decline to imports, prices and fuel. The federal environmental account of the zone and the federal economic account of the fishery most exposed to it do not reference each other. So the crew count measures exposure, not attributed harm, and the accompanying population file records not-derivable rather than a manufactured chain. Several larger groups cannot be counted at all from these sources: finfish, crab and oyster harvesters, dockside and processing employment, recreational fishing and coastal tourism work in four states, basin residents who pay to have nitrate removed from drinking water, and the farm operations inside the roughly 60 percent of the basin that yields over 80 percent of the nitrogen reaching the Gulf as of 2023.

Needs a new measurementthe number of affected people cannot be derived from what these sources publish.

Under what conditions does it arise?

1. The cost is upstream and the benefit is downstream. Nitrogen leaving a field in the Corn Belt produces its dead zone benefit on the Louisiana shelf, inside a different member state. The gain is not divisible into a share that a single acting state could claim.

2. The coordinating body cannot bind its members. Its stated role is to provide executive level direction and support for coordinating the actions of participating organizations. No source opened in this round states that membership carries an obligation, and the shared target lives in an Action Plan rather than in a rule.

3. The federal override is closed at three separate points. Section 502(14) excludes agricultural stormwater and irrigation return flows from the point source definition, section 402(l)(1) bars the Administrator from requiring such a permit or requiring a state to require one, and in 2015 the Fifth Circuit confirmed that EPA may decline to determine that federal numeric criteria are necessary so long as it explains itself.

4. The regulated half of the load moves slowly even where it can be ordered to move. In Illinois, point sources in 2020 had cut total phosphorus 16 percent against a 2011 baseline but total nitrogen only 4.7 percent. Enforceable authority over a sector is not the same thing as a large reduction from it.

5. Weather sits between the intervention and the indicator. May 2026 discharge ran about 36 percent below the 1980 to 2025 average while the nitrate load ran about 4 percent above it. A wet year and a dry year can reverse the headline number without any change in what leaves the fields.

What has been tried?

Two directions have run in parallel for a quarter century. One is money without obligation, and it is the larger of the two by orders of magnitude. The other is an attempt to convert the goal into a duty, tried twice in court and closed both times — once by a holding that the federal step is discretionary, once on state law immunity before the merits were reached.

What was found?

Why is it still unsolved?

Coordination failure — the power to act exists, it sits in twelve separate places, and nothing binds any of them to the shared number.

The first thing to establish is that the authority is not missing, because the accountable body says so itself. The 2023 Report to Congress names state nutrient reduction strategies as the cornerstone of the effort on the ground that states have the authority, with strong support from federal partners, to achieve the nutrient loss reductions needed. Read plainly, that sentence locates the lever inside each state. Nothing in federal water law stops a state from regulating farmland under its own law. The question is therefore not why the tool is absent. It is why a tool held in twelve hands is not picked up.

The answer is where the cost falls against where the benefit lands. Nitrogen leaving a field in Iowa or Illinois delivers its dead zone benefit on the Louisiana shelf, roughly a thousand river miles downstream and inside a different member state. The state that pays is not the state that gains, and the gain cannot be divided into a share the acting state could keep. No state improves its own position by moving first, and nothing makes it worse off for standing still. That asymmetry is the exact thing a coordinating body exists to bridge, and this one cannot bind anyone. Its stated role is executive level direction and support for coordinating the actions of participating organizations. The shared target lives in an Action Plan rather than in a rule, and no source opened in this round states that membership carries an obligation of any kind.

What follows is the signature of a target nobody is bound to. A target that nobody is bound to can be moved, and in 2015 it was moved by twenty years. The 2001 goal was a five-year average under 5,000 square kilometers by 2015, subject to the availability of resources; in 2015 the Task Force affirmed the goal and reset the date to 2035, a change recorded in a single sentence of its own report with no source describing a consequence to any member. The measure of the target moved as well. Two load metrics were adopted, and as of 2021 they disagreed in sign for nitrogen, one showing a 23 percent reduction and the other an increase relative to the baseline period. In January 2018 the flow normalized metric was designated the one to be used going forward. In June 2026 the interim target was declared met on that metric. Each step is defensible on its own technical merits, and the body explains flow normalization in public — that is not the point. The point is that when nobody is bound to a number, nothing in the system pushes back on the sequence.

Three legal facts belong here as evidence that the override from above is closed, not as the failure itself. Agricultural stormwater and irrigation return flows sit outside the federal point source definition, and the statute separately forbids the Administrator from requiring such a permit or requiring a state to require one. In 2015 the Fifth Circuit confirmed that EPA may decline to determine that federal numeric criteria are necessary so long as it explains itself. In 2017 the drainage district theory, which was the route to the tile drained acres where most Corn Belt nitrogen actually moves, died on state law immunity and redressability before any court reached the merits. Together these close the federal path from above. None of them removes state authority, which is why the mechanism is coordination and not exemption.

What observation would mean it is solved?

Candidates — (a) the five-year running average falls below 1,900 square miles and stays there through a full cycle (b) both load metrics show reductions approaching the 45 percent the science assessment called for, on a data record that has stabilized (c) a mechanism appears that attaches a consequence to a member state whose loads do not fall.

(a) alone is met by weather. The 2026 season shows how. A measured year of 1,332 square miles, produced by low discharge and a storm that mixed the water column, mechanically pulls a five-year running average toward the target while nothing upstream has changed. A running average can be met by weather in a year when nothing else moved. It can be lost the same way, and the forecast for that same season was 7,027 square miles.

(b) alone rests on a record that can be revised. The two metrics have already disagreed in sign for nitrogen, and the body that uses them states that flow normalized estimates are considered provisional until ten years of new data have been added. A stopping condition that can be revised backward after it is declared met is a weak stopping condition. The 45 percent figure also dates from 2007, and no source opened here restates it against current conditions.

(c) alone counts paper. States already submit management programs containing annual milestones, and no source opened identifies what follows from missing one. A new obligation written down is not the same as a load falling. The three have to be read together, and the first has to be read against the discharge year it came from.

What is it connected to?

Fills with researchnitrate removal costs borne by basin drinking water utilities, harmful algal blooms in inland lakes fed by the same loads, commodity and crop insurance policy as a driver of cropping patterns, wetland restoration and drainage water management as delivery pathways, and hypoxic zones outside the United States. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• Whether the 28 percent is provisional. The 2023 Report to Congress states that flow normalized estimates are considered provisional until ten years of new data have been added, and the June 2026 declaration rests on 2025 data. No source opened restates that caveat in connection with the announcement, or says whether it applies.
• The current value of the other metric. The Task Force adopted two load metrics and, as of 2021, they disagreed in sign for nitrogen. The 2026 announcement publishes only the flow normalized one. No source opened states the 2025 five-year moving average for either nutrient, so the disagreement can be neither confirmed nor retired.
• What follows from missing phosphorus. EPA states that the interim target for total phosphorus has not been met, and explains it by legacy storage in soils, groundwater and sediments that can release phosphorus months or years later. No source names a new deadline, a remedy, a responsible party or any consequence to anyone. The sentence has no second half.
• Why the statutory report is not there. The 2014 amendments require a biennial Report to Congress, and the Task Force uses it in place of Action Plan reassessments. The EPA index of those reports ends at 2023. As of 2026-08-08 the announcement that the interim target was met exists as a press release rather than as the report the statute names, and no source opened explains the gap.
• Any consequence for a state that misses. No source opened in this round identifies a penalty, a sanction or any consequence for a state whose nonpoint loads do not fall — not the statute text retrieved, not the Report to Congress, not the Task Force pages, not the nonpoint source pages. The absence is consistent across every source opened, which is why it is recorded here rather than reported as a finding.
• Any link between the zone and the fishery most exposed to it. The March 2026 federal economic report on the Gulf shrimp fishery contains zero occurrences of hypoxia, dead zone or oxygen. No source opened attributes any share of any population loss to the extent of the zone, which is why the affected population is recorded as not derivable.
• What full implementation would cost, or what funds it after FY2026. No source opened states what implementing the twelve state strategies would cost. Iowa alone identified USD 512 million in one reporting period against its own 45 percent goal, and the entire federal Gulf Hypoxia Program is USD 60 million over five years, with no source placing those two numbers beside each other. That program is authorized only through FY26, which is the current year; only a local news outlet notes that future authorization is uncertain, and no federal source opened addresses it.

See the evidence

Two primary documents were read end to end and they are the spine of this dossier — the 2023 Report to Congress, which is the self-published account of the accountable body, and the 2026-08-05 measurement announcement from LSU and LUMCON. Everything about the goal, the baseline, the two metrics, the deadline extension and the 45 percent science figure comes from the first. The 2026 measured value and the storm explanation come from the second. Federal agency pages supplied the membership, the goal framework, the program funding and the report index; statutory text came from a law school code repository; two case summaries came from a court opinion service and a university agricultural law center rather than from the dockets themselves. The single weakest link is the pair of headline figures the June 2026 announcement rests on. The 28 percent and 13 percent were read only in an EPA press release characterizing USGS data; the USGS landing page for that data publishes no figures, and the near real time USGS service returned HTTP 403, so those two numbers are not independently verified here and are graded low for that reason. Disagreements are left visible rather than resolved. EPA reports nitrogen down 28 percent as of 2025 while LSU and LUMCON state eight weeks later that no significant nitrate reduction has occurred since 2001, and neither document acknowledges the other or notes that the two measures are not comparable. The 2023 report gives flow normalized nitrogen as down 23 percent as of 2021 against the 28 percent as of 2025, which is a different year rather than a contradiction. The long-term average zone size is 5,223 square miles over 39 years in one source and 5,200 over 40 years in another. The 2017 record is over 22,000 square kilometers in one and 8,800 square miles in another. Even the name of the water body differs by source and by year — the NOAA and EPA documents of 2025 and 2026 say Gulf of America, while the 2026 measurement announcement and the United States Code say Gulf of Mexico — and it is left as each source has it. This is a Path A output, so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range Not derivable

The reason and what is missing are listed under “What is missing” below