Universal Child Care

Cap what a family pays for child care at seven percent of its income, pay nothing below three quarters of the state median, and fund pre-school places for every three- and four-year-old.

AI evaluation · not yet reviewed by a human

This evaluation was produced and sourced by an AI model; a human review is still pending. Figures and conclusions may still change. The review log is at the foot of the page.How review works →

Child care in the United States costs 13,184 dollars a year on average, more than in-state college tuition in most states, and public help reaches about one eligible child in six. The bill before Congress turns the subsidy into an entitlement: families pay a sliding share of their income capped at seven percent, families below three quarters of the state median pay nothing, providers are paid a rate that covers the cost of meeting quality standards, and states are funded to offer pre-school to every three- and four-year-old. It leaves delivery where it is — centres, family homes and schools all qualify — and pays for it out of general revenue rather than a contribution. This evaluation looks fifteen years ahead, because the effects on the children only appear at that distance.

Balance

Balanced · 0.53 previous scale

Balance on the previous scale. The Bilanz 2.0 simulation is not yet available for this evaluation. The category comes from the share of the debate on the pro side (r).

For 25 · 53 % Against 22 · 47 %

Size class: large Scale of this evaluation: Normalised Impact — unitless, calibrated to this topic. For comparison: one point here is worth roughly 10 billion euro per year. Two pieces of evidence pull in opposite directions and they are not about the same thing. Boston ran a lottery for its pre-school places and followed the children for twenty years: those who won were substantially more likely to finish school and reach college. Quebec made full-time care cheap for children from infancy and followed those cohorts too: they were more aggressive at school and, as young adults, more likely to be charged with a crime. This evaluation treats the first as evidence about three- and four-year-olds in school-based places and the second as evidence about babies and toddlers in full-time care, which is the reading that lets both stand — read either as evidence about the whole programme and the result moves in that direction. How we score →

Arguments for

Arguments against

7 arguments evaluated · Scoring v1.3 Δ absolute +3

Arguments — For

4 arguments · top 3 shown

What families stop paying

15of 100

The average price of child care is 13,184 dollars a year, about a tenth of a two-parent income and a third of a single parent's. Under the cap most of that bill moves from the family to the federal budget.

Value 5 · Household budgetsImpact 5.3Plausibility 5.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money families no longer pay, priced at the middle of the scale like all money. The federal budget pays it instead, which is the argument standing against this one, and the difference between the two is what a euro is worth in a household paying a third of its income for care against what it is worth in the treasury. Only the part of the federal outlay that replaces a bill families are already paying is counted here; the rest buys care that does not exist today, and what that care does is counted in the two arguments after this one rather than as money. Nothing is counted for the relief of not having to choose between a job and a child, which is the same fact as the money. Nothing is counted for the child care workers whose pay the higher rates would raise, which is a further transfer and is not estimated here. The value is the middle of the scale, and only the part of the public money that replaces a bill families already pay appears in this argument.

No official estimate exists for the bill as introduced in 2025. Estimates for its relatives run from about 60 billion dollars a year for the 2017 version of the same bill to more than 190 billion for a full entitlement covering every child under five ^[3], and the Congressional Budget Office put the 2021 child-care-and-preschool package, the closest costed design, at 381.5 billion dollars over ten years with a slow phase-in ^[9]. Eighty-five billion a year is used here as a setting, or 73.3 billion euro, in a range from 60 to 190. Not all of that lands as relief: the total American bill for paid child care is smaller than the programme, because the programme also raises what providers are paid and brings children into care who are not in it today. Forty-five percent is taken as relief of bills families already pay, in a range from 30 to 60 percent, giving 33.0 billion euro. Those families are spread across the income distribution but weighted below the middle of it, since the cap binds hardest where income is lowest; a weight of 1.6 is used, in a range from 1.2 to 2.1. That gives 52.8 billion euro a year. The Impact is the largest on this side and it is deliberately only part of the public money, because the rest buys something rather than replacing something.

▸ Show calculation ▾ Hide calculation

	Federal cost of the entitlement Setting, range 60 to 190 billion dollars: no estimate exists for the 2025 bill; the lower figure is the 2017 version of the same bill, the upper a full entitlement for every child under five, and the 2021 package was costed at 381.5 billion over ten years [3][9]	85 billion dollars a year	73.3 billion euro a year
×	Share that replaces a bill families already pay Setting, range 30 to 60 percent: the rest raises what providers are paid and brings children into care who are not in it today [4]	45 %	33 billion euro a year
×	Weight of a euro in the households the cap binds hardest for Setting, range 1.2 to 2.1: care costs a tenth of a two-parent income and a third of a single parent's	1.6	52.76 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	5.28

Score 5.28 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 15 of 100

What the programme costs is a projection and the range around it is wide, which is unavoidable for an entitlement whose cost depends on how many families take it up. The counterfactual is the present subsidy, which reaches about one eligible child in six. The chain is named — cap on family payment, federal reimbursement of the difference, lower household bill — and the link carrying the quantity is take-up, which is behavioural: an entitlement that requires an application will not reach everyone entitled to it, as the present system demonstrates. The confounder that matters is the split between relief and new spending, which no source measures directly and which is set here rather than found; a programme that mostly raises provider rates would deliver much less relief than assumed. That is named and unresolved. Reverse causation does not arise. The Plausibility is a little above the middle: the arithmetic of a cap is simple and the two numbers it runs on, cost and the share that is relief, are both estimated.

evidence basis: Projection · P ceiling 6 identification: Mechanistic · rung ceiling 6

Counterfactual: the present subsidy system, which reaches about one eligible child in six. Design: mechanistic — the cap arithmetic is statutory, but take-up and the split between relief and new spending carry the quantity and neither is measured. Confounder: a programme that mostly raises provider rates would deliver far less household relief than assumed; named and unresolved. Direction: no reverse causation. Ceiling: projection 6.0 binds and mechanistic gives the same.

Children who go further

5of 100

Boston allocated its pre-school places by lottery and followed the children for two decades. Those who won were more likely to finish high school, to sit the college entrance test and to enrol, and less likely to be locked up as juveniles.

Value 8 · Life chancesImpact 1.1Plausibility 6

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is what happens to a child's life afterwards: finishing school, sitting the entrance test, enrolling in college, and not being locked up at sixteen. This site places life chances a class above ordinary money, because what is at stake is the range of lives a person can go on to have rather than the size of a bank balance. Later earnings are used as the measuring rod, which is the only unit anyone has followed for twenty years, but they are the measurement and not the good. Read as ordinary money at the middle of the scale, this argument would be a third smaller — that is the one place where the value class rather than the arithmetic moves the result. Nothing is counted for the juvenile incarcerations avoided, which the same study measures and which would enlarge this. The value is a class above money because the stream is the range of lives available to a child, with later earnings standing in as the only long measurement anyone has.

Gray-Lobe, Pathak and Walters followed more than 4,000 children through seven Boston pre-school lotteries and found that winning a place raised college enrolment by 8.3 percentage points, alongside higher high-school graduation and lower juvenile incarceration ^[5]. The programme here would keep about 1.5 million additional three- and four-year-olds in publicly funded pre-school at any time, on top of the roughly 2 million already in state pre-kindergarten or Head Start; since each child spends two years there, about 750,000 children a year enter who would not have. Applying the measured effect gives about 62,000 additional people entering college in each year's cohort. What that is worth over a lifetime is set at 120,000 euro each, in a range from 50,000 to 250,000: enrolling is not graduating, and the premium is earned only by those who finish. Weighting by where these families sit gives 1.4. That is 10.5 billion euro a year, counting one entering cohort a year. The Impact is a fifth of the relief above and it is the only argument here whose effect arrives fifteen years after the money is spent.

▸ Show calculation ▾ Hide calculation

	Additional three- and four-year-olds entering publicly funded pre-school each year on top of about 2 million already in state pre-kindergarten or Head Start; a place is occupied for two years, so the entering cohort is half the stock [1]	1.5 million additional places at any time, two years each	0.75 million children a year
×	Additional college enrolment measured on children who won a Boston pre-school lottery against children who entered the same lottery and lost [5]	8.3 percentage points	62,250 people a cohort
×	Lifetime value of enrolling Setting, range 50,000 to 250,000 euro: enrolling is not graduating, and the premium is earned by those who finish	120,000 euro each	7.47 billion euro a cohort
×	Weight of a euro in these families the children who gain most are in households below the middle of the distribution	1.4	10.46 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	1.05

Score 1.05 Impact × 8 Value × 6 Plausibility ÷ 10 = 5 of 100

The counterfactual is children who entered the same lotteries and did not get a place, which is as clean a comparison as social policy research offers. The design is experimental — random assignment by oversubscribed lottery, with the losers followed in the same administrative records as the winners — and it is the strongest evidence in this evaluation by a wide margin. The confounder that would normally matter, that families who apply differ from those who do not, is removed by construction because both groups applied. Reverse causation cannot arise from a lottery. What holds the plausibility well below what the design would allow is the transfer: Boston's places were part-day, school-based, taught by teachers paid on the school scale, and rationed enough to be oversubscribed, none of which describes a national subsidy paid to whatever provider a family chooses. A programme that funds places without reproducing what made those places work would not produce this result. The Plausibility is at the middle to upper part: the finding is as well identified as social policy gets, and the thing it identifies is not quite the thing being proposed.

evidence basis: Study · P ceiling 6.5 identification: Experimental · rung ceiling 8.5

Counterfactual: children who entered the same oversubscribed lotteries and did not get a place. Design: experimental — random assignment across seven Boston pre-school lotteries, both groups followed in the same administrative records (Gray-Lobe, Pathak and Walters) [5]. Confounder: selection into applying, removed by construction since both groups applied. Direction: reverse causation cannot arise from a lottery. Ceiling: experimental 8.5 less two points for the transfer from part-day, school-based, teacher-staffed and rationed places to a national subsidy paid to any provider a family chooses.

Parents who can take the job

4.4of 100

When Quebec made care cheap, the employment of mothers with young children rose by nearly eight percentage points and stayed there. The same constraint binds in the United States, where about a million mothers of under-fives are out of work and name child care as the reason.

Value 5 · Household budgetsImpact 1.5Plausibility 6

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is earnings that begin to arrive because somebody can take a job they could not take before, priced at the middle of the scale like all money. It is separate from the relief above in the ordinary way: a family that was already working and now pays less gains from that argument, a parent who was not working and now is gains from this one. What is counted is not the whole wage but the part that is a net addition, because the hours now spent at work were previously spent looking after the child, and that time had a value. Nothing is counted for the career effects that follow years later, which are real and would enlarge this and are not measured for this population. Nothing is counted for the employer. The value is the middle of the scale, and the time the parent gives up is taken out of the quantity rather than the value.

Baker, Gruber and Milligan found that Quebec's subsidy raised the employment of mothers with young children by 7.7 percentage points, one of the largest labour supply responses measured for any policy ^[6]. About 12 million American mothers have a child under six, and roughly a million of them are out of the labour force and name child care as the reason. The figure used here is 850,000 additional mothers in work, in a range from 400,000 to 1.6 million: below what the Quebec elasticity applied to the whole group would give, because American mothers of young children already work at higher rates than Quebec's did in 1997. Their earnings average about 33,000 euro a year. Only 40 percent of that is counted, because the hours now spent at work were hours at home before, and what a parent gave up in taking the job — that time, near the net wage for the parent at the margin — is deducted here; the care that replaces those hours is bought with the public money counted on the other side, not here. Weighting by where these households sit gives 1.3. The Impact is a fifth of the relief above, which is the honest size of a labour supply effect once the time it uses up is subtracted.

▸ Show calculation ▾ Hide calculation

	Mothers of children under six who take a job they could not take before Setting, range 400,000 to 1.6 million: below what Quebec's 7.7 point response applied to the whole group would give, because American mothers already work at higher rates [6]	of about 12 million mothers of under-fives	850,000 mothers
×	Annual earnings each [8]	33,000 euro	28.05 billion euro a year
×	Share that is a net addition the hours now at work were hours at home before; the value of that time to the parent, near the net wage at the margin, is deducted here — the replacement care is paid for by the public outlay on the other side, not here	40 %	11.22 billion euro a year
×	Weight of a euro at these incomes mothers held out of work by care costs are concentrated below the middle of the distribution	1.3	14.59 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	1.46

Score 1.46 Impact × 5 Value × 6 Plausibility ÷ 10 = 4.4 of 100

The counterfactual is the rest of Canada, where no comparable subsidy existed, which is what the study's difference-in-differences design uses across the years around Quebec's 1997 introduction. That design is quasi-experimental, the policy change was sharp, and the labour supply result has been replicated on the same data by later authors who disagree with the paper about everything else. The confounder that matters is that Quebec's economy and its family policy moved together in those years — the subsidy arrived alongside other family measures — and the paper addresses it by finding the effect concentrated among mothers of children in the eligible age range and absent for mothers of older children. Reverse causation is not plausible: mothers' employment rates do not cause a provincial subsidy. What is unresolved is the transfer, and it is large: Quebec started from much lower maternal employment than the United States has today, so the same subsidy has less room to work here, which is why the figure used is well below what the elasticity alone would give. The Plausibility is at the middle to upper part: the effect is measured on a clean design in a country whose starting point was not this one.

evidence basis: Study · P ceiling 7 identification: Quasi-experimental · rung ceiling 8

Counterfactual: the rest of Canada, where no comparable subsidy existed. Design: quasi-experimental — difference-in-differences around Quebec's 1997 introduction (Baker, Gruber and Milligan) [6]. Confounder: other Quebec family measures arriving in the same years; addressed by the effect appearing only for mothers of children in the eligible age range. Direction: reverse causation implausible, maternal employment does not cause a provincial subsidy. Ceiling: quasi-experimental 8.0 less one point for the transfer from a province with much lower maternal employment than the United States has today.

Where the higher price goes

0.8of 100

A price that rises because places are scarce does not vanish: it is paid to the people who run and staff those places. Child care workers are among the lowest-paid in the country, so a euro that moves to them counts for more than the euro the family above the line gave up.

Value 5 · Household budgetsImpact 0.4Plausibility 4.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money arriving with child care providers and, through them, with their staff, priced at the middle of the scale like all money. It exists so that the price rise on the other side is not counted as if the money burned: a higher price on an existing place is a movement of money from the family that pays it to the provider that receives it, and the difference between what a euro is worth at each end is the whole of the welfare effect. Child care workers earn around the twentieth percentile of American wages, so their share weighs more than the family's euro; the owners' share weighs less. Nothing is counted here for any new place or any better care, which would be real resources rather than a movement of money and belong to the outlay argument opposite. The value is the middle of the scale, and this argument is the receiving end of the price rise counted against the measure.

The price rise on the other side moves about 3.2 billion euro a year from families above the eligibility line to the providers of the roughly 4 million places they use. Where it lands within the sector is set rather than measured: labour is the largest cost in child care and the bill ties public rates to staff pay, so 60 percent is taken to reach workers, in a range from 40 to 80 percent, and the rest stays with owners as margin. A child care worker earns around 30,000 dollars a year, which places the household in the second fifth of incomes, where a euro counts 1.5; owners are taken at the fourth fifth, where it counts 0.7. The blended weight is about 1.2, in a range from 1.0 to 1.4, giving 3.8 billion euro a year. The quantity is the same as on the paying side by construction, and so is the doubt about whether the price rise happens at all. The Impact is a little larger than the price rise it mirrors, because the money moves from households above the middle to workers below it.

▸ Show calculation ▾ Hide calculation

	Higher price paid on existing places by families above the cap	from the third argument against the measure	3.2 billion euro a year
×	Share reaching child care workers as pay, weighted where it lands Setting, range 40 to 80 percent: labour is the largest cost in child care and the bill ties public rates to staff pay; workers at about 30,000 dollars a year sit in the second fifth of incomes [1][4]	60 % × 1.5	2.88 billion euro a year
+	Share staying with owners as margin, weighted where it lands owners are taken at the fourth fifth of incomes; together a blended weight of about 1.18, range 1.0 to 1.4	40 % × 0.7	3.78 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	0.38

Score 0.38 Impact × 5 Value × 4.5 Plausibility ÷ 10 = 0.8 of 100

This argument stands or falls with the price rise it mirrors: if places do not become scarce and prices do not move, neither side of this movement of money exists, so the plausibility is the same as there. The counterfactual is the same market without a public entitlement. The chain is complete — a higher price is paid, the provider receives it, part of it reaches staff — and the only link that is this evaluation's own is the split between wages and margin, which no source measures for this market. The counter-mechanism is the same as opposite: the bill's supply measures may prevent the price rise altogether, in which case nothing moves. Reverse causation does not arise. The Plausibility is below the middle for the same reason as on the paying side: the mechanism is standard and nothing has measured it here.

evidence basis: Mechanism · P ceiling 6 identification: Mechanistic · rung ceiling 6 band: Chain closed, unevidenced · P 4–5

Counterfactual: the same market without a public entitlement. Design: mechanistic — the receiving leg of the price rise booked as con-3, same quantity and same plausibility by construction; the wage-margin split is set. Confounder: the bill's provider-rate and workforce measures, which could cancel the price rise entirely; named and unresolved, shared with con-3. Direction: no reverse causation. Ceiling: mechanistic 6.0 binds. Band: chain closed but unevidenced — links named, counter-mechanism stated, only the size missing.

Nothing measured argues against the claim; what is missing is any measurement of the price rise for this market and of how it splits between wages and margin. The counter-mechanism — that the bill's supply measures cancel the rise — is named and unresolved. Read back: about half the time, a price rise of the size assumed opposite occurs and reaches providers and their staff roughly as assumed here.

Open: Provider price and wage data before and after the programme starts, matched to the share of subsidised places in each market, would show both whether prices move and where the money goes, and could carry this and the paying side above 6 together.

Arguments — Against

3 arguments

Eighty-five billion dollars a year

20of 100

An entitlement of this shape costs somewhere between 60 and 190 billion dollars a year, paid out of general revenue with no contribution behind it; no official estimate exists for the 2025 bill itself. It is the largest single outlay in any evaluation on this site.

Value 5 · Public financesImpact 7.3Plausibility 5.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is federal money spent, priced at the middle of the scale as public money always is. The whole outlay appears here, not only the part that relieves family bills, because the treasury pays all of it; the part that buys new care appears on the other side as the care itself rather than as money, which is how a public euro is traced to what it actually purchases. Nothing is added for the taxes that would have to rise or the borrowing that would have to happen, which is the same euro under a later name. Nothing is deducted for the income tax the newly employed mothers would pay, which is small beside the outlay and would properly belong with their earnings. Whether the money is found by taxing, borrowing or cutting makes no difference to the weight. The value is the middle of the scale, and the whole outlay stands here while what it buys stands opposite.

The figure used is 85 billion dollars a year, or 73.3 billion euro, a setting inside a range that runs from about 60 billion for the 2017 version of this bill to more than 190 billion for a full entitlement covering every child under five ^[3]; the 2021 package, the closest design that has been officially costed, came to 381.5 billion dollars over ten years with a phase-in ^[9]. The width of that range is not a failure of estimation; it is the difference between capping what families pay and funding a place for every child, and the bill sits between them. What makes the cost hard to pin down is that the reimbursement rate is defined by the cost of meeting quality standards rather than by a dollar figure, so the price of the programme is whatever it turns out to cost to staff it. Public money carries the standard weight of 1.0. Against the 33 billion of it that relieves family bills, the remaining 40 billion buys care that does not exist today, and whether that is money well spent is the question the two arguments opposite answer. The Impact is the largest single figure on this site, and half of it is not a transfer to anybody but the price of buying care that is not being bought today.

▸ Show calculation ▾ Hide calculation

	Federal cost of the entitlement Setting: no estimate exists for the 2025 bill; the lower figure is the 2017 version of it, the upper a full entitlement for every child under five [3][9]	85 billion dollars a year, range 60 to 190	85 billion dollars a year
÷	In euro exchange rate used throughout this evaluation	1.16 dollars to the euro	73.28 billion euro a year
×	Weight of a euro in the federal budget the standard weight this site uses for public money	1.0	73.28 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	7.33

Score 7.33 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 20 of 100

The direction and the order of magnitude are not in doubt; the level is. The counterfactual is the present system of block grants and tax credits, which spends about 35 billion dollars a year in total. The chain from statute to outlay is short — entitlement, reimbursement rate, number of children — but the rate itself is defined in the bill by reference to the cost of quality rather than by a number, which makes the outlay depend on what states set and on what providers charge once they are guaranteed payment. That is a behavioural link and it is the reason this is not a definitional cost. The confounder that would push the figure up is provider pricing: an entitlement that pays whatever quality costs invites the cost of quality to rise, which is the mechanism the third argument on this side describes. Reverse causation does not arise. The Plausibility is a little above the middle: that the programme costs tens of billions is certain, and where in a three-to-one range it lands depends on rates nobody has set yet.

evidence basis: Projection · P ceiling 6 identification: Mechanistic · rung ceiling 6

Counterfactual: the present block grants and tax credits, about 35 billion dollars a year. Design: mechanistic — the entitlement structure is statutory but the reimbursement rate is defined by the cost of meeting quality standards rather than by a figure, so provider and state behaviour carry the quantity. Confounder: provider pricing under a guaranteed payment, which would push the outlay up; named and unresolved. Direction: no reverse causation. Ceiling: projection 6.0 binds and mechanistic gives the same.

What Quebec found

1.5of 100

Quebec's cohorts of children who grew up with cheap full-time care from infancy were more aggressive at school and, as young adults, 19 percent more likely to be charged with a crime and 22 percent more likely to be convicted.

Value 9.5 · Health and safetyImpact 0.3Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is the children themselves: more anxiety and aggression at school age, worse self-reported health and life satisfaction as young adults, and more of them charged and convicted. It sits in the class this site uses for life, health and protection from crime, one step below the top because none of what is measured is death. The victims of the additional offences are inside the same stream rather than counted separately, since the criminal justice figures are the measurement of both. Nothing is counted for the parents, whose gains stand opposite. Nothing is counted for the children who would have been in informal care of worse quality and are now in a regulated place, which is the counter-effect and would reduce this. The value sits just below the maximum: the stream mixes the health of children with the harm done by the offences some of them commit.

The programme would keep roughly 1.2 million additional children under three in centre-based care at any time, which is the part of it Quebec's experience speaks to; since a child spends up to three years in that age band, about 400,000 children a year newly enter it, or 11 in every 100 of the 3.6 million born each year. Baker, Gruber and Milligan's long-run follow-up of the Quebec cohorts found criminal accusations up 19 percent against the average rate, alongside worse self-reported health and life satisfaction ^[7]. That is an effect measured on whole birth cohorts, in which the use of care rose by about 15 in every 100 children ^[6]; it is carried over here in proportion, so an American cohort in which 11 in every 100 newly use care gets three quarters of the Quebec increase. Applied to a lifetime probability of being charged of about 15 percent across the 3.6 million births, that is roughly 77,600 additional people accused per cohort, at 30,000 euro each once the offence, the process and the effect on the person's own life are counted, in a range from 10,000 to 80,000. The health and life-satisfaction findings are put at 0.05 quality-adjusted years for each child who newly enters care, in a range from 0.01 to 0.2 — a fortnight of good health, which is a low reading of an effect the authors describe as substantial. Together that is 3.13 billion euro a year. Nothing is subtracted for the children moving out of unregulated care, which would cut this and is not measured. The Impact is a twentieth of the money in this debate, which is small for a finding this uncomfortable, and it is small because the effects are modest changes in probability across a large cohort.

▸ Show calculation ▾ Hide calculation

	Children born each year	3.6 million births	3.6 million children a year
×	Additional people charged with an offence, at the Quebec cohort effect measured on the Quebec cohorts followed into administrative crime records [7]	15 percent lifetime probability, 19 percent higher	102,600 people a cohort
×	Scaled to the change in care use 1.2 million additional children under three at any time, up to three years each, is 400,000 entering a year out of 3.6 million births; the Quebec effect is a whole-cohort effect of a rise in care use of about 15 percentage points [6][7]	11 in 100 newly in care here ÷ about 15 in 100 in Quebec	77,600 people a cohort
×	Cost of one such case Setting, range 10,000 to 80,000 euro: the offence, the process and the effect on the person's own life	30,000 euro	2.33 billion euro a year
+	Worse health and life satisfaction among the children who newly enter care Setting, range 0.01 to 0.2 years: a fortnight of good health, a low reading of an effect the authors describe as substantial [7]	400,000 children at 0.05 quality-adjusted years each, 40,000 euro a year	3.13 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	0.31

Score 0.31 Impact × 9.5 Value × 5 Plausibility ÷ 10 = 1.5 of 100

The counterfactual is the rest of Canada across the same birth cohorts, which is the design the authors use throughout their Quebec work. It is quasi-experimental, the policy change was sharp, and the long-run paper follows the same cohorts into administrative crime and health records rather than relying on recall. The confounder that matters is that Quebec differs from the rest of Canada in more than child care, and the paper addresses it by comparing cohorts within Quebec that were and were not eligible by birth date as well as across provinces. Reverse causation does not arise. The reason the plausibility sits at the middle rather than higher is the transfer, and it is a serious one: Quebec's programme was famously short of regulated supply, pushed large numbers of infants into whatever places existed, and paid providers too little to staff them well — while the bill here funds providers at the cost of meeting quality standards and leaves parents free to keep a child at home. A finding about a badly supplied programme is not automatically a finding about a well supplied one. The Plausibility is at the middle: the effect is well identified in Quebec and what it identifies may be the quality of the care rather than the fact of it.

evidence basis: Study · P ceiling 6 identification: Quasi-experimental · rung ceiling 8

Counterfactual: the rest of Canada across the same birth cohorts, and within Quebec the cohorts made ineligible by birth date. Design: quasi-experimental — difference-in-differences on the 1997 introduction, cohorts followed into administrative crime and health records (Baker, Gruber and Milligan) [7]. Confounder: Quebec differs from the rest of Canada in more than child care; addressed by the within-Quebec eligibility comparison. Direction: no reverse causation. Ceiling: quasi-experimental 8.0 less two points for the transfer from a programme short of regulated supply and paying providers too little to one funded at the cost of meeting quality standards.

Prices rise where places cannot

0.6of 100

Child care is short of staff before any subsidy arrives, and a guaranteed payment for a place that cannot be created raises its price instead. Families just above the eligibility line pay that price and get nothing back.

Value 5 · Household budgetsImpact 0.3Plausibility 4.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money leaving households that the programme does not cover, priced at the middle of the scale like all money. The households concerned sit above the eligibility line, so a euro there weighs slightly less than the average one, which is why this argument is smaller than its face value. The same money arrives with providers and their staff, many of them paid close to the minimum wage, and it stands on the other side as its own argument rather than being netted here; read together, the two show a price rise that moves money towards lower incomes rather than one that burns it. Nothing is counted for any new place or better care the higher price might buy, which would be a real resource and belongs with the outlay. Nothing is counted for families who cannot find a place at all, which is the present situation and not a change. The value is the middle of the scale, and the money this price rise moves to care workers is booked on the other side, where it lands.

About 4 million children whose families would be above the cap are in paid care today. The mechanism is the ordinary one for a subsidy in a market that cannot expand quickly: places are limited by staff, staff are limited by pay, and a guaranteed public payment bids for the places that exist before it creates new ones. Eight percent is used as the price increase for unsubsidised families, in a range from two to twenty percent, on an average bill of about 10,000 euro. That is 3.2 billion euro a year, at a weight of 0.8 for households above the eligibility line; the same 3.2 billion arrives with the providers and their staff, and is counted where it lands as the fourth argument for the measure. The effect is temporary in principle and the length of the transition is what the range is about: at the low end supply responds within a few years, at the high end the staffing shortage that exists today persists and the money mostly moves prices. The Impact is the smallest here, and it would be larger in the first years of the programme than in its tenth.

▸ Show calculation ▾ Hide calculation

	Children in paid care whose families are above the cap [4]		4 million children
×	Average annual bill [4]	10,000 euro	40 billion euro a year
×	Price increase for families outside the programme Setting, range 2 to 20 percent: at the low end supply responds within a few years, at the high end today's staffing shortage persists and the money moves prices	8 %	3.2 billion euro a year
×	Weight of a euro above the eligibility line these households sit above three quarters of the state median income	0.8	2.56 billion euro a year
÷	Normalised Impact scale of this evaluation	10 billion euro a point	0.26

Score 0.26 Impact × 5 Value × 4.5 Plausibility ÷ 10 = 0.6 of 100

The counterfactual is the same market without a public entitlement. The chain is complete and each link is ordinary economics: guaranteed payment, demand at a fixed supply, higher clearing price for those outside the programme. What is missing is any measurement of the size for this market. The counter-mechanism is real and named: the bill pays providers a rate tied to the cost of quality and funds workforce measures, which is a supply response rather than pure demand, and a programme that succeeds in raising the number of staff would not raise prices at all. That is unresolved. Studies of college aid capture find pass-through of roughly a quarter, which is the closest analogy available and is not this market. Reverse causation does not arise. Because nothing measures the effect and a genuine counter-mechanism sits against it, the number rests on a complete chain rather than on evidence. The Plausibility is below the middle: the mechanism is standard, nothing has measured it here, and the bill's own supply measures may cancel it.

evidence basis: Mechanism · P ceiling 6 identification: Mechanistic · rung ceiling 6 band: Chain closed, unevidenced · P 4–5

Counterfactual: the same market without a public entitlement. Design: mechanistic — chain named (guaranteed payment, fixed short-run supply, higher clearing price outside the programme) with no measurement for this market; the nearest analogy is capture of college aid at about a quarter. Confounder: the bill's own provider-rate and workforce measures, which are a supply response and could cancel the effect entirely; named and unresolved. Direction: no reverse causation. Ceiling: mechanistic 6.0 binds. Band: chain closed but unevidenced — links named, counter-mechanism stated, only the size missing.

Nothing measured argues against the claim; what is missing is any estimate of subsidy capture in the American child care market. The counter-mechanism — the bill's own provider-rate and workforce funding — is named and unresolved. Read back: about half the time, families above the eligibility line pay roughly the increase assumed here.

Open: States that have expanded child care assistance publish provider rate schedules, and market-rate surveys record what unsubsidised families pay. Comparing the two across states before and after an expansion would measure the capture directly and could carry this above 6.

Summary

This is the most expensive measure evaluated on this site and it comes out ahead by a narrow margin. What families stop paying is the largest single item, and it is worth more than the same money in the treasury because a care bill of 13,184 dollars falls hardest on the households least able to carry it. The second and third items are the ones that make the case interesting: about 850,000 mothers who could take a job they cannot take now, and a lottery study from Boston showing that children who got a pre-school place were 8.3 percentage points more likely to reach college two decades later. Against that stands the outlay itself — for which no official estimate of the 2025 bill exists — and a warning from Quebec, where cheap full-time care from infancy produced cohorts who were more aggressive at school and more likely to be charged with a crime as adults. The result therefore depends on reading Boston as evidence about three- and four-year-olds and Quebec as evidence about babies — take either as evidence about the whole programme and the balance moves in that direction.

Outlook — effect over time

Balanced · 0.53 previous scale

Sources

Congress.gov: H.R. 4418, Child Care for Working Families Act of 2025. congress.gov
Center for American Progress: What to know about the Child Care for Working Families Act. americanprogress.org
U.S. House Committee on the Budget: Universal child care. budget.house.gov
Child Care Aware of America: Child care in America: price and supply. info.childcareaware.org
Gray-Lobe, Pathak and Walters, Quarterly Journal of Economics: The Long-Term Effects of Universal Preschool in Boston. academic.oup.com
Baker, Gruber and Milligan, Journal of Political Economy: Universal Child Care, Maternal Labor Supply, and Family Well-Being. nber.org
Baker, Gruber and Milligan, American Economic Journal: Economic Policy: The Long-Run Impacts of a Universal Child Care Program. aeaweb.org
U.S. Census Bureau and Bureau of Labor Statistics: Families with own children: employment status of parents. bls.gov
Congressional Budget Office: Economic Effects of Expanding Subsidized Child Care and Providing Universal Preschool (H.R. 5376, November 2021): 381.5 billion dollars over 2022 to 2031. cbo.gov

Last reviewed by Claude Opus 5 · September 14, 2026 · 2× AI, not yet reviewed by a human

September 14, 2026AI reviewClaude Opus 5re-scored
Gesamtprüfung 08.09.: Bestände in Jahreskohorten umgerechnet (pro-3 i 20,9→10,5), Quebec-Effekt über Nutzungsanteil skaliert (con-2 i 3,43→3,13), Empfängerbein pro-4 zum Mehrpreis neu (i 3,78), Kostenquelle als Setzung + CBO 2021 ergänzt; r 0,56→0,53.
September 6, 2026AI reviewClaude Opus 5First evaluation
First evaluation: Boston lottery read as evidence about pre-school and Quebec as evidence about care from infancy, which is what lets both stand.

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →