Thread by @lymanstoneky

I have a piece out today on childcare policy. I'll review it in a moment but I want to start on a less policy, more personal note. ifstudies.org/blog/more-choi…

Over the course of writing about this issue, there's been a lot of misunderstanding about the debate on childcare, and indeed people have a lot of different ideas about what the debate "is about."

For many people, childcare is really "about" work, and especially maternal labor force participation. I have a very hard time on a personal level accepting this thesis because it doesn't match my experience.

Among the households I know enough about to guess at this stuff when I was growing up, the households with the most working hours were not the households with the most daycare usage.

Nor was it the case in my personal experience that use of childcare necessarily implied maternal employment, nor did maternal employment necessarily imply heavy usage of daycare.

Statistically, the correlation between maternal employment and use of childcare is pretty high, but the exception cases still numbers in the millions, so it's not like "care user without employed mother" or "employed mother but non-care-user" are marginal experiences.

Throw a rock in a playground or a daycare center and after a couple tries you'll hit one of these mismatch cases. They're just not that rare. So I'm skeptical of the idea that we should mostly frame the discussion around maternal work.

Relatedly, I've had a non-trivial number of people in my mentions at various times, sometimes politely and sometimes less so, suggest that my views on childcare are shaped by my own interests and culture: that I'm just trying to subsidize my own lifestyle.

The only problem with this is that both of my parents had thriving careers for my entire life, and I spent a non-trivial amount of time at daycare. I don't know how much time because I was little, but I have plenty of daycare memories.

Daycare was fine for me, and the daycare I attended is well-regarded. I have no reason to think it had any "adverse effects". However I would be lying if I didn't say that every *actual memory* I have of daycare is unpleasant. Of course that's just the selection bias of memory!

There's one particularly horrifying one involve rocks and my nose but I won't regale you with the details.

But my point is, my desire to provide at-home care with equal treatment is not motivated by an actual cultural background of opposition to daycare, nor have I ever been a member of a family where only one spouse had a busy career.

For Ruth and I, I quit my very cushy Federal job with a 6-figure paycheck so we could move to Hong Kong for a vocation we both believed in, but which also happened to be her career. WIth Suzie born, she is continuing that career.

Suzie is not in daycare even though here in Quebec it is free, because although we both have busy careers, we have spent years making painful and costly decisions with huge income penalties in order to ensure we both work flexible jobs compatible with childcare.

Even so we're not philosophically opposed to putting Suzie in daycare for some length of her life. Ruth and I are both products of Kentucky public schools and if we can find one that is suitably flirtatious with the limits of constitutionally allowable religion, we'd enroll Suzie

(the schools I grew up in were public but often sooooo flirtatious with the constitutional limits that, well, it was maybe more than flirting)

However, we just really like raising Suzie at home with us. We value it. We made costly sacrifices to make it happen. We didn't make this decision because it's how we were raised; both of us had two-career parents who did excellent jobs raising us.

We made the decision because it's what we decided was best for us and our child. And while I do not expect the government to subsidize my decision over and against someone else's decision, I absolutely demand that the government not *discriminate* against it.

The government of Quebec forthrightly and explicitly discriminates against my decision. One major reason they do so is that I speak English and they really really REALLY want Suzie enrolled in a French daycare.

Of course, I would actually LOVE for Suzie to learn French. However, the kids at the daycares who show up at our local parks have a tendency to not always be very nice to Suzie, so I'm not chomping at the bit for them to be her peers.

All that to say, when we think about childcare policy, I sincerely believe it is a critical mistake to think about it primarily in terms of "parental work arrangements," and think it's much better to think about it in terms of parental values and child welfare.

So in my article, I do not address labor force effects of childcare programs. I don't even identify them as a factor worth contemplating for such a policy, because I sincerely think that organizing children's lives around the convenience of parental work is a bad idea.

If parents choose to make this arrangement that's fine! I trust them! But we should not use the (inherently coercive!) power of the state to choose one specific such arrangement to the disadvantage of others.

Okay, now, to the article....

.... except not really, because I have SPECIAL BONUS CONTENT ONLY FOR TWITTER! whaaaat!

A key argument I make in the article is that childcare usage is disproportionately for the well-to-do, and that simply making it free won't change that. I show some cross sectional evidence across states with different programs as evidence.

But while I think it's pretty clear, you might disagree. Maybe you'd rather see "What happens in an actual case of expanding free childcare?"

We have a great case study of this! In 2014, Vermont rolled out free universal early childhood education.

So for example, here's the difference in school enrollment rates by child age between Vermont and New Hampshire, by year. What you can see is after 2014, enrollment rates for 3 and 4 year olds in VT rise WAY above NH.

So making it free DOES boost enrollment! The effect is about 10-20%, which coincidentally is the exact same difference we observe in state-level cross-sectional data.

What happened to the income mix of attendees?

No change. The structural inequality in participation was *identical* pre- and post-reform.

Note that the ACS sample sizes in VErmont are small; that's what drives the noisy estimates.

But the clever reader will spot a problem!

Maybe free childcare doesn't change the income ratio ***because it enables formerly nonworking parents to work***!

We can test this by comparing the average income from wages and salaries among parents with children under 5 in VT vs. NH (and alternatively comparing young-child parents in VT to older-child parents in VT).

The result?

MAYBE a small positive change? But nothing obvious?

So probably there was soooooome work effect. But it's evidently not that big, and appears very noisy.

And if we look at **total household income**, which would account for spousal earnings, non-wage income, welfare, tax credits, etc, we also see no change. So to the extant labor income rose, other income fell.

So what we can say is that when Vermont expanded free childcare:
1) Childcare enrollment rose by a large amount
2) The income gap between enrollees and non-enrollees was unchanged
3) Enrollee moms maybe had a SMALL work increase
4) But it was offset by other income decreases

This doesn't mean that universal childcare is a bad policy!

Tons of parents want childcare, it's stupidly expensive, helping those parents out is a fine thing to do.

But we should not selectively only help the parents who want to use center-based care.

We should help all parents. Providing an at-home care per-child allowance worth somewhere between 50-75% of the per-child subsidy for center-based care is entirely reasonable and is more consistent with supporting parental choice.

The children cared for at home are uniquely likely to be low-income or ethnic or linguistic minorities, i.e. people who the state might have a uniquely strong interest in helping get a hand up in life, ergo, at-home care allowances will tend to be highly progressive.

(This by the way is why some conservatives oppose at-home care allowances, because they correctly understand they disproportionately go to poorer families and families with fewer working hours)

I'll extend the thread to go through the article itself later on this morning. Have some other stuff to do for a bit now.

In the article, I argue four key considerations should drive childcare policy: choice, compensation, fairness, and cost. These boil down to "spend money giving parents more choices about childcare in a way that doesn't discriminate/ accelerate inflation." ifstudies.org/blog/more-choi…

On choice, I show that there are huge numbers of nonusers of institutional care who are probably worth helping. The work parents (especially mothers!) do caring for kids is valuable and just as worth supporting as center-based care.

The government has no *compelling interest* in encouraging *specifically* center-based care. The government has an interest in defraying the escalating costs of childcare, and in strengthening parental choice of care. So a home-care allowance is important.

Since many of the costs of childcare have nothing to do with tuition prices but rather are about opportunity costs (for parents who provide at-home care) or search costs and commute costs, disregarding non-tuition costs disregards a large share of costs.

You might say "we don't do this for public school!"

True. Two big response to that:
1) American public education is not super awesome and affording more choice in it, especially for poorer families, would be a good thing
2) Childcare/ECE =/= Primary/Secondary school

The reason we have public education historically, and the reason it is generally *mandatory* to have some kind of education, is not to provide childcare, but because there is wide societal agreement that all children need to know a certain curricula of knowledge and skills.

That curricula includes many things that parents are often not very good at teaching. Many parents discovered this very acutely during COVID! Even at the 3rd grade, teaching the stuff kids learn can be way more complicated and specialized than it looks!

TEachers do difficult and often quite sophisticated work even at the primary school level!

But virtually everything a 0-4 year old learns is entirely within the capacity of parents to teach, and indeed since most learning at those ages is by emulation and osmosis not formal instruction, parents are actually *uniquely positioned* to do the teaching.

Plus, a lot of the stuff we want kids to learn at a young age is explicitly *not* universal. To the extent we value diversity and pluralism (I value it a lot), we want parents to be able to pass on their language, culture, religion, ethnic identity, etc, to their kids.

This happens often at a very early age. Indeed, MOST of what is happening at early ages is euphemistically called "socialization," which basically means "cultural assimilation."

*gestures wildly at Quebec*

Thus, the compelling arguments for public education do not apply to early childhood care and education: there is no universally agreed upon curriculum, and parents are entirely qualified and capable of carrying out the necessary tasks.

So EVEN IF you think that choice in public education is bad, the argument for overtly supporting choice in early childhood is very strong.

My next argument is about compensation. I'm gonna skip the details and just say: childcare workers are probably underpaid, but the reasons why they are unpaid are waaaaay harder to address than advocates tend to admit.

And the most underpaid childcare workers of all are....

Parents.

If you calculate Biden's child allowance as a parenting-wage, it shakes out to between 1/6 and 2/3 as much pay per hour of parental care as what childcare workers get per child-hour of care.

But of course, care-using parents get that benefit too, so it's not really comparable that way.

I am not arguing that we should actually pay parents as hourly workers, but I'm saying IF we are going to subsidize childcare, THEN we should not erase the labor of at-home care, and should compensate it at the same-per-child-labor-subsidy-rate as center-based care.

My preference of course is to not do either thing and just to use child allowances for all parents.

Now, conservatives often hate this proposal because it creates a "work disincentive." It actually doesn't, it just balances out the work-family equation by taking the government's thumb off the scale.

(or, rather, applying the government's thumb equally to both sides of the scale)

But progressives often hate this proposal because it's subsidizing parents (especially moms) to stay home.

To which I can only respond with someone who has been very influential in my thinking: Silvia Federici!

I don't agree with Federici on everything, but she's to my mind very obviously one of the most clear-headed and thoughtful thinkers on these issues of anybody on the left.

My next argument is about fairness.

Cui bono?

Richer people.

I belabored this very heavily further up the thread so I won't hammer it too much here but folks the evidence in the US and in numerous international contexts is that the poorest families do not use these programs.

And finally, costs. This is one I think has seriously not been addressed even halfway decently by any childcare advocates.

The cost of childcare is exploding at an alarming rate, much faster than inflation.

The best estimates suggest that spending per hour of care rose 30% in inflation adjusted terms between 1990 and 2011, and obviously more since. This is driven overwhelmingly by higher-income families at for-profit daycares.

Any subsidy program for childcare that doesn't tackle the underlying cost problem will create an incredibly rapid pace of cost increases. We desperately need to tackle the cost side of the equation.

I do not have all the answers to this, but as I note in the article it's not ***just*** Baumol's cost disease. Childcare is not only labor-intensive but also *land* intensive and heavily regulated. It's a perfect storm of structural cost drivers.

I don't have a perfect list of solutions for this issue. But I'm flagging it because we really, really need more thoughtful engagement on how to deliver good care at a better price.

(I also happen to know that @PTBwrites is about to publish an article giving an excellent rundown on the cost of childcare and what to do it so consider this me bumping it for him to spike)

(or setting, or whatever, I don't volley the ball except embarrassingly badly at the beach)

Finally, I want to talk about another reason why I don't think that free childcare will radically close income gaps in who enrolls, and to the extent it does so the outcomes will not be good. And that's because... we have an experience of a big increase in childcare spending!

From 1990 to the early 2000s, the environment of childcare costs as they related to returns to work changed RADICALLY for low-income parents. Starting in 1990 a big new childcare subsidy program was rolled out, and in the mid-1990s "welfare reform" began.

So two things happened: we increased subsidies for childcare, and we increased the financial returns to work

So here's spending under the CDCTC (mostly to higher-earning households) and the federal CCDF (mostly to lower-earning households) since 1990, represented as a share of CES-estimated expenditures on daycare and nurseries + a share of CCDF expenses since...

some share of CCDF expenditures doesn't go to reimburse costs paid and reported in CES, but rather to programmatic supports, etc

Now, this isn't a full accounting of childcare costs because some CDCTC expenditures would go to forms of care that aren't reported in the CES category I use. And indeed, there are probably a LOT of childcare costs not reported in this category.

But even if you assume that I'm only capturing HALF of childcare spending, the reality is I'm also not including *any* state and local spending on early childhood education, so the total subsidy position is about the same.

The government is *already paying for* more than half the cost of childcare in the country.

But you can see that in the 1990s, there's a big change. Spending on the CCDF program grows.

Now, there was an earlier program 1990-1993 that is consolidated into CCDF in 1994. I could not find data on that. But if I just flatline the nominal spending in 1994 back to 1993....

WHOOPS, sorry grabbed the wrong column for CDCTC in the graph above! Doesn't radically change conclusions. Here it is with corrected CDCTC, and using a flat CCDF number pre-1994.

If instead you assume a linear increase from 0 in 1990, it's this, which is what I'm gonna be working from.

However, it's worth noting that these funds aren't just "funds for childcare." They are often tied to work. CDCTC is only for working parents, and CCDF often has work connections too.

So what these programs actually did was not "subsidize childcare," but "provide a higher return to work." They aren't a test of "what happens when you make childcare free" they're a test of "what happens when you subsidize work more."

(there's some fuzziness on this with CCDF in some cases but the idea generally holds)

Starting in the mid 1990s, you get another round of changes with welfare reform. In 1993 you get a big EITC expansion that directly increases returns to work, in 1996 you get PRWORA that creates benefit limits and work rules, there's a lot of state activity, etc

We know what happened to employment rates for impacted parents: they rose. A lot. There's debate about how much can be attributed to what factors, but I'm of the school of thought that welfare limits, work requirements, and the EITC probably all mattered.

So employment rose a *lot* particularly among single parents. Welfare receipt declined too. The program was a success along its intended metrics. I don't dispute any of that.

However, one of the *intended* consequences of the policy changes 1990-2000 was that single parents should put their kids in daycare at higher rates. That was an overt goal, in order to facilitate employment.

That goal was successful. Here's the chart from the article:

As you can see, the gap between married and single parents was small in the late 70s or early 80s, then grew. Then in the 1990s, the gap closes. For 4 year olds it starts to close in 1989, for 3s in 1994.

And in absolute terms, increases in childcare enrollments persist through the early 2000s.

This is precisely the timeline of welfare reform, and also the timeline of the generosity of work-tied childcare subsidies.

So, the workist reforms of the 1990s successfully pushed a huge number of kids into daycare via subsidies that, although aimed at childcare, primarily operated view *increasing returns to work*.

This matters because it helps us understand why e.g. Vermont's reform did NOT change the composition of care-users: because "free childcare" doesn't draw in kids from poor families, but "work got more appealing" it turns out very much does.

We can also ask what happened to those kids.

Did the children of welfare reform turn out okay?

In short, no. Much as the children of Quebec's childcare reform ended up with much higher crime rates, the children of welfare reform had worse behavioral outcomes.

Some exemplary links here:
nber.org/papers/w28077
nber.org/papers/w25527
dspace.mit.edu/bitstream/hand…

The point is, free childcare doesn't draw in cost-constrained people, because cost-constrained people also have really bad work options so staying home with baby *is just more appealing*. Improving work options however does boost care enrollment....

.... but the result is not that the kids in question are saved from poverty, but rather that their disadvantage in many cases *intensifies*.

Which is to say, I am genuinely confused by this reply, since I explicitly mention the 1990s childcare subsidy changes in the article, and as shown above am not ignorant of them. I think they prove my point.

https://twitter.com/AngelaRachidi/status/1395408797636628485

I am not arguing that it's impossible to narrow the gap. I'm arguing that doing so requires explicitly getting parents to trade off parenting effort for work effort, with negative consequences for their children and thus plausibly intergenerational mobility.