Joe Biden is set to release his American Family Plan later today. The plan contains a large number of items. Here and in my coming advocacy, I will focus on the child allowance, paid leave, and childcare, as those are the three keystones of family benefits policy across the world.
The earlier American Recovery Plan created a temporary two-tier child allowance implemented through the Child Tax Credit (CTC) and the Earned Income Tax Credit (EITC).
Under this temporary scheme, families receive a CTC of $3,600 for children below the age of 6 and $3,000 for children between the ages of 6 and 17. These benefits do not phase in based on earnings, meaning that even the poorest families in the country are eligible for them. These payments are supposed to be made monthly by the IRS.
Families will also receive an EITC equal to as much as $3,618 per child paid as a lump sum in the subsequent year. Families with earnings below $10,640 per year are not eligible for the full amount and the poorest families are not eligible for any EITC benefit. These payments are supposed to be made as an annual lump sum by the IRS.
There was some hope that, when it came time to extend these temporary benefits into a permanent scheme, Biden would modify these parameters a bit, especially when it comes to the EITC parameters that are intentionally designed to exclude poor children from the benefits. But he has not.
The AFP proposes to simply extend all of this to 2025. I don’t know of any other child allowance scheme in the world that deprives the poorest children of the full benefit. This scheme deprives them of more than half of the full benefit. Needless to say, this is a moral obscenity.
After year 2025, the CTC portion of the child allowance, which the poorest families are eligible for, is set to decline from $3,000/$3,600 to $1,000. The EITC portion, which the poorest families are not eligible for, is not.
The official reason for scheduling steep CTC cuts to happen in 2026 is that it allows the Democrats to play games with the CBO’s ten-year scoring window. So they don’t actually want the benefit cut to happen in 2026. They are just happy to play chicken with it under what could be a different president, different congress, or both.
Weirdly this desire to play chicken does not apply to scheduling EITC cuts in 2026, just to scheduling cuts to the only cash program that the poorest families in the country are categorically eligible for.
When it comes to paid leave, the text of the AFP is quite vague.
It will guarantee twelve weeks of paid parental, family, and personal illness/safe leave by year 10 of the program, and also ensure workers get three days of bereavement leave per year starting in year one. The program will provide workers up to $4,000 a month, with a minimum of two-thirds of average weekly wages replaced, rising to 80 percent for the lowest wage workers. We estimate this program will cost $225 billion over a decade.
There is no discussion of work history requirements, no discussion of when people will be eligible for their 12 weeks of benefits (does “by year 10” mean not until then?), and no discussion of the minimum benefit level. It doesn’t say whether this will be a fully federal program, e.g. one run by the Social Security Administration, or whether this will be a hodgepodge of subsidies to employers and states to run paid leave programs.
Some of the details listed in this paragraph mirror what is in the FAMILY Act, but some of them do not. I would guess that Biden intends to do something like the FAMILY Act, with these modifications, but until there is more clarity on that front, I will hold off on making criticisms based on that assumption.
There is one clear criticism that this sparse text invites. The plan says that the lowest wage workers will receive 80 percent income replacement. I do not think that is enough to allow many of the lowest wage workers to actually use the benefit. A minimum wage worker likely cannot give up 20 percent of their pay for 12 weeks. At that part of the wage scale, the replacement rate should be 100 percent or more.
The AFP continues the bizarre Democratic position that pre-kindergarten for ages 3 and 4 should be free but that child care for ages 1 and 2 should not. Nobody has ever explained why this is the case. It just seems to be a strange artifact of different people being assigned the two tasks.
The idea of providing free pre-kindergarten is a good one, though Biden’s plan does not provide any significant details:
President Biden is calling for a national partnership with states to offer free, high quality, accessible, and inclusive preschool to all three-and four-year-olds, benefitting five million children and saving the average family $13,000, when fully implemented. This historic $200 billion investment in America’s future will prioritize high-need areas and enable communities and families to choose the settings that work best for them
If you do some quick math on this, you’ll see that 5 million children multiplied by $13,000 multiplied by ten years equals $650 billion. Yet the price tag is listed here as $200 billion.
This suggests that Biden is either planning to implement the pre-kindergarten many years into the future or trying to make states foot a large share of the bill. The first approach is a bummer and the latter approach is a bad idea on the merits that will also ensure that many especially conservative states don’t participate in the free pre-kindergarten program.
When it comes to child care for children below the age of 3, Biden initially proposes Obamacare-style subsidies that reduce the price of child care to a percentage of a family’s income:
Families will pay only a portion of their income based on a sliding scale. For the most hard-pressed working families, child care costs for their young children would be fully covered and families earning 1.5 times their state median income will pay no more than 7 percent of their income for all children under age five.
This is similar to the Murray and Warren proposals and has all of the problems those proposals have.
The income test will create administrative burdens that are hard for some families to navigate and, depending on how it is implemented, could lead to punishing clawbacks. Setting the subsidy equal to a percentage of income means that parents who space their children out will pay more towards child care than parents who have their children closer together. The user fees will discourage utilization of child care, which is the opposite of we want, while also encouraging affluent parents to put their children in segregated centers with only other unsubsidized affluent parents.
What’s funny about these kinds of proposals is that the user fees that are set equal to a percentage of a family’s income are functionally equivalent to a tax. But if I were to propose increasing taxes on families with incomes of 1.5x the median by 7 percentage points, people would call that punishing and a non-starter. Yet this is what the Biden proposal does.
These user fees are bad on the merits and don’t raise any significant revenue. They should be scrapped altogether, as is done in the pre-kindergarten proposal.
In addition to the sliding-scale income subsidies, Biden also weirdly includes a significant expansion of the Child and Dependent Care Tax Credit (CDCTC). Under this expansion, tax units can receive a credit equal to 50 percent of their child care expenses up to $4k for a one-child family and $8k for a multi-child family.
Analyzed on its own, this design will heavily exclude poor people. Many poor people cannot afford 50 percent cost-sharing for child care expenses and many more cannot afford to front 100 percent of the cost of child care and then wait as much as 16 months to get 50 percent of it reimbursed.
But the weird thing about this CDCTC proposal is how it is supposed to interact with the sliding-scale child care subsidies mentioned above. If low and moderate income people are getting subsidized upfront, then presumably they won’t have very many child care expenses to claim towards the CDCTC. So is the point of the CDCTC to provide a credit to higher-income families who do not receive much from the sliding-scale child care subsidies? In that case, why wouldn’t you just bake that into the sliding scale rather than create a separate benefit? And why would you phase out the CDCTC at $125,000 of income?
The whole thing gets even stranger when you realize that, in addition to the sliding-scale subsidies and the CDCTC, there is also the Dependent Care Flexible Savings Account (DCFSA), which was doubled to over $10,000 in the American Recovery Plan. How do these three things interact? It looks like you would get the sliding-scale subsidies first, then apply DCFSA money to any bill that still exists, and then for anything you pay beyond that, you will get 50% of that bill back in the form of a tax credit, provided your income is not too high.
Who on earth would ever design a child care benefit system like this?
Using a DCFSA to pay for child care expenses provides a tax subsidy equal to your marginal tax rate while using the CDCTC provides you a tax subsidy equal to 50 percent. Thus, it seems that only people who are affluent enough to be beyond the CDCTC phaseout would benefit from the DCFSA program because anyone with income below that has a marginal tax rate below 50 percent.
So if you piece it all together, the poorest would benefit only from the sliding-scale subsidies. As you got richer, sliding-scale subsidies would decline but some CDCTC subsidy would kick in. And as you got richer still, sliding-scale subsidies and CDCTC subsidies would decline but DCFSA would kick in. This will create any number of horizontal asymmetries and paperwork hassles as you’ll need to be an accountant to figure out which bucket to use.
Finally, all of the child care benefits provide no support to parents who prefer to do child care in the home. Parents with young children who do not take advantage of child care services and therefore are not eligible for the sliding-scale subsidies, CDCTC, or DCFSA should receive a home child care allowance instead.
If we are going to do child care benefits, let’s do them in a serious unified way that looks like we actually have some theory of what we are trying to accomplish and not just extend every duplicative vestigial program that exists, no matter how ill-fitting and strange its interactions with the overall scheme is.