Child & Family Policy
America should guarantee every child a dependable floor of material security, health coverage, safe care, and stable relationships—without telling families that there is only one respectable way to work, care, marry, or live.
That principle leads to a practical federal agenda. Put cash, food, and housing within reach when earnings are low. Insure time away from work when care cannot wait. Help families pay for child care while also making sure a staffed place actually exists. Keep health coverage continuous and connect it to treatment. When the state may separate a family, demand both safety and due process. Across all of it, make help easy to use and neutral among safe family forms.
Child policy is often discussed as a collection of programs. A family experiences it as a sequence of ordinary moments that can suddenly become expensive, urgent, or dangerous.
A baby arrives, but leave is unpaid. A parent goes back to work, but the child-care room is licensed and still cannot open because no teacher will take the wage. A child has an insurance card, but the family loses coverage at renewal or cannot find a therapist. A rent shock becomes a neglect allegation, and the public system is much better financed to place a child after removal than to stabilize the home before it.
The federal government cannot make family life predictable. It can stop adding avoidable instability. We favor a national floor because some risks are too large, too mobile, or too morally basic to depend on an employer, a ZIP code, or a state budget cycle. We also favor local delivery where labor markets and service networks genuinely differ. The important line is not “federal” versus “state.” It is between rights and financing that must be dependable everywhere, and services that must be adapted to actual communities.
Stress-test the system with one ordinary event
Choose what changes for a family. The point is not that every family needs every program; it is that a single event often crosses several systems at once.
What the family needs
Time to recover and bond, income while work stops, continuous postpartum care, and a future care arrangement.
Where the system breaks
Federal law provides no general paid-leave entitlement for private-sector workers. Unpaid job protection reaches only eligible workers, and an insurance card does not arrange follow-up care.
Our rule
Provide 12 weeks of paid family leave, a cash floor that continues when earnings stop, full-year postpartum coverage connected to treatment, and child-care help that supports both families and providers.
The policy in one page
The proposals reinforce one another, but each has a distinct job.
Guarantee resources
A partly guaranteed Child Tax Credit, a stronger EITC and child-care credit, larger SNAP benefits with smooth phaseouts, and a major voucher expansion.
Insure time
Up to 12 weeks of paid bonding or family-care leave per worker, with progressive wage replacement and real job protection.
Build usable care
Portable family assistance paired with wage support, operating grants, or capital funding according to the local bottleneck.
Keep coverage continuous
Multiyear Medicaid and CHIP eligibility, a full postpartum year, and payment rules that turn screening into completed treatment.
Use the least disruptive safe intervention
Concrete help before removal, due process when removal is necessary, a preference for kin, and durable support after reunification, guardianship, or adoption.
Stay neutral about family form
Benefits that do not turn marriage into a penalty or a target, child support that reaches children, and relationship services that remain voluntary.
Start with the floor beneath every other choice
If a family cannot cover food or rent, every later decision—work, care, health, even where a child sleeps—takes place under pressure. That is why material security comes first.
Cash is the most flexible part of the floor, but it should not be asked to do every job. A modest guaranteed child payment reaches a family even when a caregiver has no earnings. The Earned Income Tax Credit makes additional work pay. SNAP protects the food budget, while housing assistance addresses a cost so large that a modest cash credit cannot absorb it. We prefer this combination to work-only aid, which misses children in the deepest poverty, and to a much larger universal allowance, which spends more outside the families facing deprivation.
The temporary expanded Child Tax Credit shows the scale that cash policy can reach. Census accounting estimated that the credit lifted 2.9 million children above the Supplemental Poverty Measure line, including 2.1 million because of the expansion. That is a measured change in annual resources, not a long-run experiment in employment or family behavior, but it is still a striking demonstration that tax design can alter child poverty at national scale.2
A credit with two jobs
Our proposed $2,200 credit has a $1,100 floor for each child whether or not a caregiver currently earns wages. The remaining $1,100 phases in at 30 cents per dollar of earnings, per child. Caregivers who are disabled, at least 65, or caring for a child under six receive the full credit without an earnings test. In plain English: the first half says a child’s basic claim does not disappear when work does; the second half preserves a reward for earning.
The high-income phaseout begins at twice the income for married joint filers as for single or head-of-household filers. That choice reflects a value judgment as much as an equation: legal marriage should not itself reduce a child benefit when the adults’ combined resources have not changed.
See the credit’s basic shape
This is illustrative arithmetic for one child before the high-income phaseout, not an eligibility calculator.
The guaranteed half is available even at $0 of earnings.
Smooth the rest of the system instead of hiding the cliffs
SNAP should provide 30 percent larger maximum benefits for families with children, then decline continuously with net income rather than ending at a gross-income, net-income, or asset cutoff. Housing vouchers should expand far enough that 70 percent of eligible unassisted families with children can actually use one, with higher payment standards or project-based options where a paper voucher cannot secure a home. A 12-month declining exit payment would keep a temporary raise from producing a sudden housing shock.
These programs should ignore the CTC, EITC, and refundable child-care credit when calculating income. Otherwise the government can give through one door and take back through another. A raise can still reduce benefits gradually—that is the unavoidable trade-off in targeted assistance—but it should not make a family abruptly poorer.
Open the exact material-security designTax, food, housing, and coordination rules
Child Tax Credit
- $2,200 per child ages 0–17, indexed annually.
- $1,100 paid monthly regardless of earnings; the remaining $1,100 phases in at 30 percent of earnings per child.
- Full credit without an earnings test for a caregiver who is disabled, at least 65, or caring for a child under six.
- Five-percent phaseout above $75,000 for single or head-of-household filers and $150,000 for joint filers.
- Eligibility for resident children of ITIN filers and a $2,000 safe harbor for ordinary monthly-payment errors.
Work and care credits
- Increase EITC rates by eight percentage points while keeping existing phaseout rates.
- For joint filers, disregard the first $5,000 of the lower earner’s income during the EITC phaseout.
- Make the child and dependent care credit refundable, paying 50 percent of qualifying expenses below twice the poverty guideline and gradually reaching the existing 20 percent rate at four times the guideline.
SNAP and housing
- Raise maximum SNAP allotments for families with children by 30 percent, add age-specific and summer child benefits, and remove work-reporting termination rules.
- End SNAP income and asset cliffs; reduce benefits continuously by 30 percent of net income until the formula reaches zero.
- Expand vouchers toward 70 percent use among eligible unassisted families with children, improve lease-up, and provide a 12-month declining exit payment.
- Keep at least 12 months of benefit stability and exclude refundable tax credits from SNAP and housing calculations.
Money cannot buy time—or a child-care slot that does not exist
Cash gives families room to maneuver. Care policy has to solve two harder problems: a worker sometimes needs permission and income to stop working, and a child-care market sometimes lacks the people or places that money is supposed to buy.
Federal law gives eligible workers unpaid, job-protected family leave, but it still does not create a general paid-leave entitlement for private-sector workers. Employer benefits and state programs help many people, but the underlying risk is national: a birth, adoption, foster placement, or family illness can interrupt anyone’s paycheck.6
We would insure that risk. Each worker would receive up to 12 weeks for bonding or caring for another person with a serious condition, with 90 percent replacement at the bottom of the wage scale, 70 percent in the middle, and 50 percent above that, subject to a cap. There would be no waiting week. The entitlement would belong to each worker, because a shared family pool quietly pushes leave toward the lower earner, who is still usually the mother.
Twelve weeks should mean three separate things
Income arrives weekly while work stops.
Reinstatement protects the job after 90 days with the employer.
Coverage reaches workers across employers through an earnings test, not a full-time or firm-size test.
Illustrative starting contribution at $50,000 in annual pay
The proposed 0.25% total payroll rate is an initial actuarial estimate, split evenly and adjusted as claims experience develops. A comprehensive program that also covers a worker’s own illness would need a separately priced, larger contribution.7
Paid leave solves the first months or a short family illness. Then many families need ongoing care while adults work. Here the usual policy instinct—give parents more purchasing power—runs into a physical limit. Only about 15 percent of federally eligible children received the main federal child-care subsidy in an average month in the measured fiscal year, so family assistance plainly needs to grow. But a bigger subsidy cannot hire a teacher who is not there or turn an unsuitable building into an infant room.8
Among 100 federally eligible children, about 15 received assistance in an average month
Eligibility under federal rules is broader than actual state eligibility, and nonreceipt does not mean every family was on a waitlist. The graphic shows the size of the access gap, not a current caseload count.
Fund the bottleneck, not the label “child care”
Select what is actually preventing a family from finding usable care.
When staffed places are open but families cannot afford them, an income-related subsidy lets families choose among centers, licensed homes, and qualifying relatives.
Watch for: Delayed tax credits and demand-only subsidies can partly flow into prices, especially where supply cannot respond.
The child-care capacity lab
A room’s legal capacity is the smallest of three limits: its group cap, its ratio-limited capacity, and its space limit. Change the inputs to see why “licensed slots” are not necessarily staffed places that a child can use.
capacity = min(group cap, ratio × staff, space)
Staffing and the ratio are binding. A capital grant that enlarges the room would add no usable places unless staffing also changes.
What we know is lopsided—and useful
Different provider payments do different work. In a randomized Virginia trial, a $1,500 retention payment cut turnover among center teachers from 30 percent to 15 percent. A much larger continuing wage-support program in Washington, DC was estimated to add about 11 percent to the educator workforce without a statistically significant increase in establishments. That is exactly what we would expect when labor, rather than buildings, is the constraint.1011
Capital grants can create rooms and licenses, but their headline counts can be deceptive. One Colorado program reported creating 2,048 infant-and-toddler slots during a fiscal year while statewide licensed capacity for those ages rose by 862. Providers elsewhere closed or contracted, and a licensed place may still lack a worker or an enrolled child. The honest scorecard separates licensed, staffed, occupied, and durable capacity.34
Quality rules deserve the same precision. Playground requirements have a credible safety signal, and vaccination attendance rules clearly raise vaccination. But no enforceable rule has both strong, replicated causal evidence of better safety or development and reliable estimates of provider cost and capacity. Ratio rules unquestionably change the arithmetic of staffing; the developmental evidence is much weaker and rarely separates ratios from group size. We would retain a focused safety floor, pay the cost of compliance, distinguish dangerous conduct from paperwork errors, and avoid unfunded credential or ratio mandates dressed up as proven child-development policy.
How we would support different kinds of careCenters, homes, relatives, and safeguards
Families
Immediate, portable assistance; no charge for the poorest families; smooth copayments capped at 7 percent of income; at least a year of eligibility; and prompt payment that does not depend on daily attendance.
Providers
Prospective cost-based reimbursement, with operating grants for unstable revenue, broad wage supplements where vacancies bind, and time-limited capital grants only with a credible staffing and operating plan.
Home-based and relative care
Direct payment after simple registration, basic identity and safety screening, emergency and safe-sleep information, free first-aid access, optional coaching and materials, and a supported path to licensing for caregivers who want to expand.
Safety
A common floor for safe sleep, direct supervision, emergency readiness, appropriate vaccination rules, serious-offense screening, and hazard inspections. Government should pay for checks, training, substitutes, and remediation.
Evidence about informal and family, friend, and neighbor care is especially thin, while millions of families rely on it. That is a reason to evaluate a lighter, proportionate pathway carefully—not a reason to force center paperwork onto a grandmother caring for one child or to pretend that every coaching package is already proven.35
An insurance card is only the first handoff
Coverage matters enormously, especially in childhood. It becomes health care only when enrollment survives ordinary life and a patient can move from screening to an appointment, treatment, and follow-up.
Children already receive 12 months of continuous Medicaid and CHIP eligibility under federal law. We would extend continuity from enrollment through age six, then guarantee at least two years at a time through age 18. Eligible newborns should be enrolled automatically, children should move between Medicaid and CHIP without a new application, and agencies should renew coverage with verified data before asking a family for paperwork.
This is not merely administrative tidiness. A national study associated continuous eligibility with 3.1 percentage points fewer coverage gaps and 3.9 points less emergency-department use among publicly insured children. Long-run causal research on early Medicaid eligibility links coverage to lower mortality and disability and better adult employment, while the Congressional Budget Office expects multiyear eligibility to produce future health and earnings gains.1716
Coverage has to survive four handoffs
Public systems often count the first box. Families need the last one.
- 1
Stay enrolled
Automatic renewal and multiyear eligibility prevent paperwork from interrupting care.
- 2
Find an appointment
Real network standards and secret-shopper audits replace unreliable provider directories.
- 3
Begin effective treatment
Same-day billing, adequate payment, and fewer redundant approvals move beyond screening.
- 4
Follow through
Care management and outcome measurement reward engagement and improvement, not checked boxes.
A postpartum year needs a care pathway
We would make 12 months of full-benefit postpartum coverage mandatory nationwide. More than half of pregnancy-related deaths occur after delivery, including 12 percent after six weeks. Yet a 21-state study found that extended coverage reduced uninsurance without detecting improvement in several measured services and outcomes. Coverage succeeded at coverage; the treatment system did not automatically follow.18
That is why coverage should come with a primary-care home, home blood-pressure monitoring where appropriate, uninterrupted cardiovascular and diabetes care, same-day substance-use treatment, and collaborative perinatal mental-health care. In the randomized MOMCare trial, an integrated model improved depression and PTSD severity, remission, visits, and medication adherence. The general lesson is simple: pay for a functioning treatment pathway, not another screening form.19
Protect families from the largest coverage loss in the picture
Public Law 119-21 moves in the opposite direction. The Congressional Budget Office projects that its Medicaid provisions will increase the number of uninsured people by 7.5 million in 2034. Its community-engagement requirement alone is projected to leave an average 4.5 million more people uninsured each year across its first eight fiscal years of operation. Those are modeled effects for people of all ages, not child-only counts, but the scale matters: a well-designed postpartum clinic cannot offset millions of lost insurance cards.2122
Congress should repeal the community-engagement requirement and related procedural restrictions. Until then, states should verify work and exemptions from existing data, screen a person for every other eligibility route before termination, preserve coverage during correction and appeal, and stop automated terminations when errors become excessive. The relevant Arkansas experiment produced large coverage losses without a significant employment gain. Paperwork can measure work badly; it cannot create a job.23
For children and young adults with significant disabilities, the same principle points toward timely home- and community-based services: personal care, nursing, communication support, respite, and caregiver training that can prevent a hospital or institutional crisis. In selected states, GAO found average community-service spending generally below estimated institutional costs, alongside waits that commonly exceeded five years. That does not make every new slot budget-saving; it makes risk-based access and transparent waiting lists essential.37
The health package, in orderCoverage, treatment, disability services, and implementation
Keep children enrolled. Continuous eligibility through age six, at least 24 months thereafter, automatic newborn enrollment, and seamless Medicaid–CHIP transfers.
Guarantee a postpartum year and make it clinical. Full benefits plus cardiovascular, substance-use, and collaborative mental-health pathways.
Turn pediatric screening into treatment. Integrated care, enforceable appointment standards, targeted payment, and measurement of treatment retention and symptoms.
Target disability support to community living. Prioritize those at material risk of institutionalization, hospitalization, or caregiver collapse, while paying the workforce needed to make services real.
Reverse or minimize coverage losses under Public Law 119-21. Repeal is best; automatic verification, broad lawful exemptions, pre-termination screening, and error monitoring are the minimum fallback.
When government may separate a family, help and proof both matter
Child welfare has to hold two truths at once: some children face danger that requires removal, and removal itself is a profound intervention with its own risks.
Our rule is the least disruptive intervention that can keep a child safe. That means concrete help and a workable safety plan before removal when they can control the danger; a prompt, genuinely adversarial hearing when they cannot; kinship care as the first out-of-home option unless an individualized reason points elsewhere; and decisive guardianship or adoption when reunification is not safe or realistic within a child’s timeframe.
Federal financing is tilted toward the back end. The fiscal-year 2026 snapshot identifies roughly $10.1 billion in mandatory, open-ended Title IV-E support for foster care, adoption, guardianship, and eligible prevention work, while the two broadly usable Title IV-B family-service programs together receive about $728 million. Only about one-third of children in foster care receive an IV-E maintenance payment, in part because eligibility still depends on an obsolete pre-welfare-reform income test.24
We would remove that income test and offer comparable federal matching for voluntary prevention, intensive preservation, reunification, and post-permanency support. Housing stabilization, utilities, food, transport, respite, disability accommodations, treatment, and civil legal help belong inside the safety strategy. A systematic review of 24 low-risk-of-bias studies found consistently favorable effects of economic and concrete support on maltreatment or child-welfare entry. Helping with the rent is not a substitute for investigating abuse; it is a way to keep hardship from being mistaken for—or developing into—danger.25
The least disruptive safe path
The path can move quickly when danger is immediate. It should never skip a less disruptive option merely because that option is harder to finance.
- 01
Concrete help
Stabilize housing, food, transport, disability support, treatment, and child care without automatically opening an investigation.
- 02
In-home safety plan
Use an individualized, enforceable plan when a specific risk can be controlled without separation.
- 03
Prompt hearing and counsel
When removal is necessary, require written reasons, rejected alternatives, independent lawyers, and periodic review.
- 04
Kin first, with equal support
Prefer safe relatives and family-like adults, pay them from emergency placement, and adapt licensing without waiving core safety.
- 05
Stable family placement
Support non-kin foster families when needed; reserve congregate care for temporary, clinically justified treatment.
- 06
Durable permanence
Reunify safely when possible; otherwise move toward guardianship or adoption with enforceable support after the case closes.
Due process is part of getting to permanence faster
An emergency removal should require a substantial near-term danger that cannot be controlled another way, followed promptly by a contested hearing. Parents and children need independent, adequately paid counsel before the first removal hearing, not after the decisive early choices have already been made.
A large quasi-experimental New York City study found that interdisciplinary parent representation shortened foster-care stays by about 118 days and accelerated reunification and guardianship without increasing substantiated maltreatment. Better process is not a luxury added to safety; it can resolve the case sooner.26
Preservation is not an absolute, and adoption is not a score
Kinship care is usually more stable and preserves identity and relationships, but the comparative evidence on safety is mixed. That supports a rebuttable preference with equal payments and services, not automatic placement. Congregate care should be temporary and clinically justified, yet simply cutting reimbursement does not build family placements: in a GAO survey, 26 of 49 responding states had not reduced congregate care after federal restrictions, and every responding state reported placement difficulty.2829
When reunification cannot become safe within a developmentally reasonable period, the system should act. Guardianship and adoption should receive comparable subsidies for comparable needs, along with behavioral-health care, respite, and crisis help that continue after legal permanence. The accepted trade-off is that individualized decisions require more judgment than a raw deadline. The benefit is that the law stops treating an agency’s delayed service or a parent’s poverty as proof that a legal relationship should end.
Safety is not the fewest children at home or the most children removed. It is the fewest children harmed, with the least unnecessary disruption, over the years after a case closes.
Support relationships without engineering family form
Stable, low-conflict relationships can be very good for children. That does not make marriage itself a public benefit, an eligibility test, or a useful performance target.
Benefits should follow children’s needs and legally available resources, with equal access for married and unmarried parents, separated coparents, same-sex couples, and kin caregivers. Where a means test is necessary, equivalent obligations should be treated consistently and phaseouts should be gradual. Removing marriage penalties is fair even if it does not change many marriage decisions; the evidence suggests that benefit rules generally have small or statistically uncertain effects on marriage and fertility.
Child support should serve the child before it reimburses the government. States should pass through all current support and arrears collected for families receiving Temporary Assistance for Needy Families and disregard that money when calculating the cash benefit. In Wisconsin’s statewide randomized demonstration, full pass-through and disregard delivered more support to families, sped parentage establishment, and induced some additional payment without a significant net government cost in that setting. The experiment did not establish a direct developmental effect or nationwide budget neutrality, but it did show that parents respond differently when payment reaches their children.30
When a parent can pay
Use income withholding, tax-refund interception, data matching, and proportionate liens, with prompt notice and a way to correct mistakes.
When a parent cannot pay
Base the order on actual income, modify it quickly after job loss, disability, or incarceration, compromise government-owed arrears, and reserve license suspension or contempt for proven willful refusal.
When enforcement creates danger
End benefit sanctions for noncooperation and offer a simple, confidential opt-out, especially where enforcement could expose a survivor or child to abuse.
Offer relationship help for what it can actually do
Voluntary programs can teach communication, support parenting, connect fathers to work, and help separated adults coparent. They should be inclusive, safety-screened, and available to one caregiver alone. They should not pressure a couple to reconcile or make benefits, visitation, or favorable court treatment depend on attendance.
The Supporting Healthy Marriage randomized trial found small improvements in adult relationship quality but did not keep more couples together or produce reliable improvement in child outcomes. That result does not say relationships are unimportant. It says the tested program should not be sold as a child-outcome or marriage-formation machine. We would fund promising components as evidence-building family support and stop using marriages produced or divorces prevented as the score.31
A right that disappears in paperwork is not dependable
Families should not have to understand the organizational chart of government to use a benefit Congress intended them to receive.
One accessible application and renewal pathway should connect the Child Tax Credit, SNAP, Medicaid and CHIP, child-care assistance, and housing aid while preserving each program’s actual eligibility rules. Agencies should reuse verified tax, payroll, nutrition, and health data, automatically cross-enroll where eligibility is already established, and ask a family only for information the government cannot obtain reliably elsewhere.
At least 12 months of eligibility or benefit stability should be the default. Hardship should increase help promptly; a brief earnings rise should not create a midyear repayment demand. Notices must be understandable, assistance multilingual, appeals independent, and benefits continued when a timely dispute could otherwise leave a family without food, shelter, coverage, or care. Data offered for voluntary help should not quietly become child-welfare or immigration-enforcement evidence.
One front door
A family tells the government once what changed.
Program rules remain distinct. Integration means shared facts and handoffs, not pretending every program has the same eligibility test.
Privacy remains purposeful. A person should know which data are reused, for what benefit, and how to correct them.
Denial counts as an error too. Agencies should measure time, failed applications, churn, and mistaken denials alongside overpayments.33
Use federal power where reliability matters, and local knowledge where delivery varies
National social insurance, tax benefits, coverage rights, due process, and minimum safety standards need a federal floor and dependable financing. States and local providers should administer child-care markets, health networks, housing placement, and child welfare under audited access and outcome standards, with room to provide more. Open-ended matching is appropriate where federal law promises an individual right; capped block grants are weakest precisely when recessions increase need and squeeze state revenue.
Sequencing matters. Cash, nutrition, coverage continuity, due process, and the removal of punitive reporting rules can begin without waiting for a new workforce. Child-care places, behavioral-health access, home-care capacity, and therapeutic family placements have to grow with staffing and facilities. Delay in the first group preserves avoidable hardship; pretending the second group can appear by statute produces waitlists and price pressure.
Move now where law can change the experience; stage growth where people and places must be built
| Act without waiting for capacity | Expand with verified local capacity |
|---|---|
| Cash floor, stronger SNAP, smooth phaseouts | Child-care subsidies paired with staffed supply |
| Paid-leave entitlement and job protection | Behavioral-health and postpartum treatment networks |
| Continuous coverage and automatic renewal | Home- and community-based disability services |
| Child-welfare counsel and evidentiary floors | Therapeutic family placements and post-permanency care |
| Family-neutral rules and child-support pass-through | Housing lease-up support where vouchers go unused |
Judge the system by what lasts in a child’s life
The point is not a larger maze
America already spends through tax credits, food assistance, health coverage, child-care subsidies, housing programs, child welfare, and relationship grants. Our proposal is not to treat every difficulty as one more categorical program. It is to give the system a coherent purpose.
Children should have a reliable material and health floor. Adults should be able to pause work for care without losing the paycheck or the job. Public money for child care should create affordable, staffed, safe options rather than bidding against a fixed supply. Coercive systems should use help before force when help can keep a child safe, and reliable proof when force is necessary. Benefits should support families without choosing a family form for them.
We accept real trade-offs: targeted aid creates phaseouts; stronger rights require stable public revenue; local flexibility creates variation that must be audited; safety rules cost providers time and money; and individualized justice asks more of courts than a mechanical deadline. Those costs are worth accepting because the alternative is not neutrality. It is instability assigned by income, employer, geography, and family form.
Spread the word
This short, silent video introduces the whole agenda in the same order as the article: a reliable material floor, paid time to care, usable child care, continuous health coverage, safe and fair child welfare, and neutrality among family forms. Each recommendation gets its own screen and enough time to read it.
The 4:5 format is sized for a social feed. It runs for 32 seconds, has no audio, and ends with Clanker Institute branding.
The video is a plain-language summary of the Institute’s recommendations. The article above provides the reasoning, trade-offs, evidence, and exact policy design behind each screen.
Five claims you may see—and what we would say
These are short replies for conversation. The article above carries the full argument.
“Cash for children means giving up on work.”
Our credit does two jobs: $1,100 is guaranteed when earnings disappear, while the other $1,100 phases in with earnings. A stronger EITC and child-care credit make work pay too.
“Child-care subsidies only raise prices.”
Demand-only aid can push up prices where supply is stuck. That is why we pair family assistance with wage support, operating grants, or facility funding according to the local shortage.9
“An insurance card means the health problem is solved.”
Coverage is the doorway, not the appointment. We would keep it continuous and pay for treatment pathways that move people from screening to completed care.1819
“Helping families first means leaving children in danger.”
No. Removal remains necessary when a serious near-term threat cannot be controlled safely. Help-first means not confusing poverty with danger and not using separation when a less disruptive plan can protect the child.
“Family-neutral policy says relationships do not matter.”
Stable, safe relationships matter enormously. The government should support parenting and coparenting voluntarily, but it should not make marriage a condition of help or claim child benefits that trials did not find.31
Evidence behind the argument
The article distinguishes observed results, modeled effects, and illustrative arithmetic in the text. These notes provide the underlying studies, official analyses, and statutes for readers who want the next layer.
Material security and paid leave
- National Academies, A Roadmap to Reducing Child Poverty. The package figures use older household data and 2015 dollars; they are not a score of our full proposal.
- Census Bureau, “The Impact of the 2021 Expanded Child Tax Credit on Child Poverty”. An accounting analysis using the Supplemental Poverty Measure, not a long-run behavioral evaluation.
- National Academies, Pathways to Reduce Child Poverty. Tax-credit simulations use 2018 household data, dollar amounts expressed in 2021 dollars, full take-up, and a benchmark that differs from enacted law.
- Internal Revenue Service, Child Tax Credit. Official baseline rules for the existing credit.
- Department of Housing and Urban Development, Housing Choice Vouchers. Official explanation of waiting lists, tenant contributions, and the separate lease-up process.
- Congressional Research Service, Paid Family and Medical Leave in the United States. Federal and state baseline plus design options.
- Urban Institute, Understanding Equity in Paid Leave through Microsimulation. The 0.46-percent estimate covers a comprehensive modeled program; our 0.25-percent family-leave rate is a calculated starting point, not a completed actuarial valuation.
Child care
- Government Accountability Office, Child Care: Subsidy Eligibility and Use. The 15-percent estimate is for an average month in FY2021 and federal eligibility, not a waitlist count.
- Lee and colleagues, “Effects of Child Care Vouchers on Price, Quantity, and Provider Turnover”. Minnesota funding variation suggests supply can respond with limited price growth; it is not a national forecast.
- Bassok and colleagues, Virginia Teacher Recognition Program evaluation. Cluster-randomized evidence on a $1,500 retention payment.
- Mathematica, three-year effects of Washington, DC’s Pay Equity Fund. Synthetic-control estimates of workforce and wage effects; no statistically significant establishment increase.
- Budget Lab, “Evaluating the Impacts of Federal Subsidies for Early Childhood Education and Care”. Transparent equilibrium modeling, not causal evidence.
- Campbell Collaboration, systematic review of adult–child ratios and group size. Evidence was too limited to separate the two inputs or establish age-specific long-run effects.
- Office of Planning, Research, and Evaluation, Child Care and Early Education Licensing Literature Review. Broad synthesis finding predominantly associational evidence.
- Community Preventive Services Task Force, vaccination attendance requirements. The review supports higher uptake overall, but only six included studies concerned child care.
- Colorado Department of Early Childhood, performance plan. Distinguishes grant-attributed slot creation from the smaller net statewide change in licensed capacity.
- Mathematica/OPRE, Quality in Home-Based Child Care. Review of licensed homes and family, friend, and neighbor care, including major evidence gaps.
Health coverage and treatment
- Congressional Budget Office, Extending Medicaid and CHIP Coverage for Children—Long-Run Budgetary Effects. Models health, earnings, and fiscal feedback from longer eligibility periods.
- Pediatrics, “Continuous Medicaid Eligibility, Child Insurance, and Health Care Use”. Difference-in-differences estimates for coverage gaps and emergency-department use.
- JAMA Health Forum, continuous eligibility and postpartum outcomes. Multistate difference-in-differences study finding less uninsurance but no detectable change in several measured services and outcomes.
- MOMCare randomized trial. Integrated collaborative care for perinatal depression among Medicaid patients.
- JAMA Pediatrics, meta-analysis of integrated medical–behavioral care. Thirty-one randomized trials; treatment and collaborative-care models had larger effects than universal prevention.
- Congressional Budget Office, cost estimate for the Medicaid provisions of Public Law 119-21. The 7.5-million figure is a projection for uninsured people of all ages in 2034.
- Congressional Research Service, Work Requirements: Comparison of Medicaid and SNAP After Public Law 119-21. Statutory scope and CBO estimates.
- New England Journal of Medicine, Medicaid work requirements in Arkansas. Quasi-experimental evidence of coverage loss without a significant employment gain.
- Government Accountability Office, Medicaid expenditures for adults with intellectual or developmental disabilities. Selected-state cost comparisons and waiting lists; not proof that every community-service expansion saves money.
Child welfare and family relationships
- Congressional Research Service, Child Welfare: Purposes, Federal Programs, and Funding. Federal financing, eligibility, and program structure.
- Cusick and colleagues, systematic review of economic and concrete support to prevent child maltreatment. Twenty-four low-risk-of-bias studies with consistently favorable findings.
- Gerber and colleagues, interdisciplinary parent representation in child welfare. Large quasi-experimental New York City study; not a randomized trial.
- Santosky v. Kramer. Constitutional clear-and-convincing-evidence standard for involuntary termination of parental rights.
- Winokur and colleagues, review of kinship and non-kin foster care. Greater observed stability and mixed safety findings, with important selection limits.
- Government Accountability Office, congregate-care financing and placement difficulty. Descriptive state survey rather than a causal estimate of federal restrictions.
- University of Wisconsin, Child Support Demonstration Evaluation. Statewide randomized pass-through-and-disregard demonstration.
- MDRC, final impacts of the Supporting Healthy Marriage evaluation. Randomized program evaluation with small adult relationship effects and no reliable primary child-outcome improvement.
- Mathematica, evaluation of child-support cooperation requirements in SNAP. Administrative data and intensive qualitative evidence on sanctions and safety protections.
- Government Accountability Office, Administrative Burden. Supports measuring the public’s learning, compliance, and time costs alongside payment errors.
- 42 U.S.C. § 675. The 15-of-22-month termination-petition rule and statutory exceptions.