The strange career of the expanded Child Tax Credit

A good idea that didn't work, and a new chance at a new strategy

Give the gift of Slow Boring this holiday season — 20 percent off on gift subscriptions from now until Christmas.

During the first year of Slow Boring’s life I wrote a lot about the Child Tax Credit, which, after a massive temporary expansion during the American Rescue Plan, succeeded in its goal of reducing child poverty — seemingly without harming labor force participation — and then expired unceremoniously.

This is a sad story. The expanded CTC was a great policy, and establishing a generous and universal-ish (more on that later) child allowance program could’ve spurred changes to the welfare state and macroeconomic management going forward. But despite high hopes, the one-year expanded CTC did not ultimately lead to a permanent program. Narrowly, that’s because Joe Manchin didn’t agree that it was a good idea. But more broadly, it’s because of a multi-faceted failure of communication and political strategy.

That failure is in need of some analysis, particularly because some CTC proponents are hoping to at least partially resurrect the program as Congress considers a year-end omnibus appropriations bill or continuing resolution.¹ The available mechanism is that various corporate tax breaks are scheduled to expire, and the belief is that Republicans (though honestly most Democrats too) will want to extend them, which potentially creates an opportunity to push for CTC changes as well.

This is, at least implicitly, a big change in strategy. It means shifting away from hoping that the CTC would be so popular that it would prove politically durable, which was a nice idea that ultimately did not work. This new strategy is much more in line with the history of other (ultimately more successful) programs like the EITC and Medicaid — fight consistently, over time, and for small incremental improvements to the program.

A bunch of annoying vocabulary and definitions

I think a lot of this stuff is underexplained in news articles, so I want to specify at the start what I mean by a few of these terms. The idea of a tax credit is that if you owe $5,000 in taxes but you’re also eligible for a $100 tax credit, the credit knocks you down to $4,900. That’s different from a $100 tax deduction, where you subtract $100 from your taxable income and the actual value of the deduction varies according to your tax rate. The upshot is that a $100 credit is worth more than a $100 deduction, but also that deductions are mostly useful to people in high tax brackets who have complicated tax situations,² and thus both actually claim deductions (most people take the standard deduction) and receive large value from them.

But what happens if you only owe $15 in income taxes and you get a $100 credit?

One possibility is that you pay $0 in income taxes, thus receiving in practice a smaller credit than richer people. But another possibility is that you receive the full $100 credit, which can be conceptualized as a credit against your payroll tax burden or against state and local taxes. That’s called a fully refundable tax credit. What’s more, Congress in its wisdom has decreed that some tax credits should be partially refundable, i.e. you get some amount that is greater than $0 but lower than the $85 over and above the $15 that you owe in income taxes.

Most tax credits are also means-tested, so their value phases out if you become too rich. That’s a way of limiting the cost of tax credit programs, though in practice it means imposing an implicit marginal tax on people somewhere in the middle of the income distribution who receive part but not all of the credit. Tax credit programs that combine partial refundability with means testing tend to have a trapezoidal structure on charts. Here, for example, is a Center on Budget and Policy Priorities chart showing how the Earned Income Tax Credit works and also showing a proposal from Sherrod Brown and Ro Khanna to make it more generous — in both cases, the poorest families get the least help (because they pay the least in taxes) and families earning around $40,000 receive assistance by also have an implicit marginal tax imposed on their earnings.

Tax credits aren’t the only kind of social program. Some programs, most importantly including Social Security but also Unemployment Insurance and TANF, pay cash benefits.

Many other programs — of which Medicare and Medicaid are the most important but also SNAP, Section 8 housing vouchers, and others — deliver in-kind benefits.

Lurking over all these programs is the specter of “welfare.” Looking at the General Social Survey or other polls, most respondents say that America spends too little on assistance to the poor but too much on welfare. Unlike the other items in bold here, welfare does not have a clear definition. For decades, the term served as an informal shorthand for the no-longer-extant program Aid to Families With Dependent Children (AFDC), which provided cash benefits with a sharp means-test to poor families. But since AFDC ended, there has been a lot of ambiguity around how to draw the line between spending money on helping the poor (good) and spending money on welfare (bad). Both Bill Clinton and Barack Obama tried to draw the following lines:

• A program is not welfare if it provides in-kind benefits.

• A program is also not welfare if eligibility requires an earnings history (Social Security, UI).

• A program is also also not welfare if it’s a tax credit with a phase-in.

But the policies that resulted left the United States with an unusually high relative child poverty rate. And Democrats’ efforts to reframe the issue didn’t deter Republicans — they have continued to try to attach work requirements to in-kind assistance programs, and the range of things that should be provided via these programs is the subject of endless debate. SNAP doesn’t cover the cost of diapers or tampons, for example, even though almost anyone would regard these items as non-frivolous necessities. It does cover soda, which is groceries but not really nutrition. You get carping from the right that recipients are buying too much steak or whatever, but because SNAP isn’t cash, you can’t save it by being unusually thrifty. Normally we’d say that someone who ate rice and lentils for months to save up to buy durable goods was engaging in wise, far-sighted behavior — but that behavior is banned by the terms of the program.

So there’s a desire on the right to categorize basically all assistance as “welfare,” and a desire on the left to kick aside the stool of welfare reform and just let people have some cash.

From American Family Act to American Rescue Plan

This move toward cash really kicked off in 2017 when Senators Michael Bennet and Sherrod Brown wrote the American Family Act, which pledged to make the maximum Child Tax Credit bigger, make it payable monthly, make it fully refundable, and start phasing it out at an income of $75,000 a year for single parents and $110,000 for married couples.

Then the 2017 Tax Cuts and Jobs Act (TCJA) passed on a party-line vote by Republicans increased the maximum CTC (though not to Bennet-Brown levels) and raised the income at which the phaseout starts. TCJA didn’t change the refundability, but by making the status quo CTC more generous (and doing so with Republican votes), it made the change to the Bennet-Brown scale cheaper. By 2019, their bill had a ton of Democratic cosponsors, and the phaseouts had been raised to $130,000 a year in income for single parents and $180,000 for married couples. The change was embraced by virtually every Democrat running for senate in the 2020 cycle.

One important exception was former vice president Joe Biden, who — sticking with the tradition of a 90s-era Democrat, which was also the tradition of the Obama administration — drew a line against cash benefits for people with no labor income. Then Biden won the nomination, which seemed like a big problem for the AFA push. But then in mid-September 2020, Biden came out in favor of an AFA-like expansion of the CTC as an emergency pandemic measure. When Biden won the election in November, and then Democrats won two runoff races in Georgia to secure narrow control of the Senate, Biden rolled out a very ambitious American Rescue Plan that included, among other things, a one-year AFA-like expansion of the CTC. As I’ve said before, I initially assumed the proposal was an anchor point for negotiation. Then when a group of 11 GOP senators showed up with a plan for $650 billion in Covid-19 relief spending, I figured moderate Democrats would bolt and insist on negotiating a deal with them that bumped the number up just slightly.

Instead, Joe Manchin, Kyrsten Sinema, and other moderates went for it and passed almost exactly the Rescue Plan that Biden had proposed.

This sent expectations soaring for a future reconciliation bill. In particular, most observers (including me) figured that everyone was aware of the context behind the expanded CTC push and agreed to do it in the Covid-19 relief bill to set the table for making it permanent down the road. After all, this was transparently not a targeted Covid-19 measure. The subsidiary hope of CTC proponents was that once it was in place, it would prove so popular that Congress in some sense “had” to extend it. But then things began to fall apart. It turned out that Manchin had fundamental objections to the idea of cash support without work requirements. It also turned out that while Brown and Bennet had a lot of cosponsors, they had not persuaded the rest of the party to pursue CTC expansion instead of other family-centric priorities, and leadership was also trying to pass subsidized childcare, preschool, and paid parental leave.

And while the higher CTC payments were popular with parents, they were modestly underwater with nonparents (which is most people), so there was very little political pressure to extend.

A series of misunderstandings

This all added up to what I think remains an underrated blunder. The CTC as a downpayment on a permanent program would have been a legacy-defining reduction in child poverty. But the CTC as a temporary program added to inflation for no real reason and generated a lot of intra-caucus strife and ill will. Manchin gets a lot of undeserved crap from the left, but I think failing to communicate his own opinions about this to the administration and to his colleagues was a big failing.