Restoration of the American Rescue Plan Act's expanded Child Tax Credit

PolicyEngine computes how the 2021 policy would affect Americans and America in 2023.

In Tuesday’s State of the Union speech, President Joe Biden called to

restore the full Child Tax Credit, which gave tens of millions of parents some breathing room and cut child poverty in half, to the lowest level in history.

Biden may have been referring to the Child Tax Credit (CTC) expansion under the American Rescue Plan Act of 2021 (ARPA), which expanded the CTC in several ways. Following our Child Tax Credit report released yesterday, this report describes ARPA’s CTC expansion, and PolicyEngine’s predicted impacts of restoring that expansion.

President Bill Clinton introduced the Child Tax Credit in 1997, and the Tax Cuts and Jobs Act of 2017 expanded it to create its current structure. For instance, in 2023, a married filer with one five-year-old child receives the following benefit from the CTC based on their earnings (view this in PolicyEngine here):

As we explain in our Child Tax Credit article, the credit phases in with income due to partial refundability and also phases out with income. Refundable credits can reduce one’s tax liability below zero, while the value of non-refundable credits cannot exceed one’s tax liability. Part of the Child Tax Credit, the Additional Child Tax Credit, is refundable, and the ACTC also phases in with income; the remainder is non-refundable.

For the 2021 tax year only, the American Rescue Plan Act made four major changes to the CTC:

1. Extended eligibility to 17-year-olds

2. Raised the maximum amount from $2,000 per child to $3,600 for children under age 6 and $3,000 for children aged 6 to 17; this bonus phases out for parents with income exceeding $75,000 (single) or $150,000 (married)

3. Made the credit fully refundable

4. Prepaid half the expected 2021 credit on a monthly basis, beginning in July 2021

This video shows how to model these changes in PolicyEngine. To see the results, skip to the next section, where the links reflect the pre-built policy.

Let’s return to the married family with a five-year-old child. Their maximum credit rises from $2,000 to $3,600. Parents with no income don’t benefit from the CTC today, so the combination of the increased credit and full refundability provides them the full $3,600. The benefit phases out to $1,600 if their income ranges from about $36,000 to $150,000, and the reform no longer provides a net benefit to households with income of about $181,000 or more.

Marginal tax rates rise in each of the income ranges where the benefit phases out.

You can estimate how reinstating the ARPA CTC expansion would affect your own household by entering your information here.

We estimate that restoring the ARPA CTC expansion would cost $100.2 billion in 2023. When scoring ARPA (2021 only), the Joint Committee on Taxation estimated a cost of $109.5 billion. OpenRG estimated $100.7 billion using the open source Tax-Calculator model (I coauthored this analysis).

While the ARPA CTC expansion provides greater benefits to low-income families than higher-income families, it provides less to low-income households than middle-income households because they have fewer children. The program also provides high-income families less because of the phase-out.

As a percentage of net income, however, restoring the ARPA CTC would strictly provide greater benefits to lower income deciles.

Restoring the ARPA CTC would benefit households comprising 41% of Americans, disproportionately those in the fourth through eighth income deciles (the current CTC benefits 45% of Americans). It would raise net income by at least 5% for 27% of Americans, disproportionately those in the fourth and fifth income deciles.

Poverty would fall by 9%, and child poverty would fall 37%. When evaluating the ARPA CTC’s impact on child poverty in 2021, the Niskanen Center and the American Enterprise Institute projected reductions of 39% and 35%, respectively.

In the State of the Union, President Biden said that the “full Child Tax Credit…cut child poverty in half”. He may have been referring to the full ARPA, which also included a round of relief checks, enhancements to the Supplemental Nutrition Assistance Program, and other reforms. The Columbia Center on Poverty and Social Policy estimated that ARPA would reduce child poverty by 56%. The Census Bureau reported that child poverty fell 46% from 2020 to 2021, from 9.7% to 5.2%.

Halving child poverty with the ARPA CTC structure would require a $4,400 benefit, rather than $3,000 or $3,600. This would cost $177 billion, 77% more than restoring the ARPA CTC expansion.

Under the ARPA CTC expansion, deep poverty — the population share below half their Supplemental Poverty Measure threshold — would fall 11%, and deep child poverty would fall 72%.

Restoring the ARPA CTC would reduce each of the three income inequality measures we report, especially the Gini index at 1.9%. The Gini index reflects low-income households more than the top 10% and 1% income shares do, and these are the groups disproportionately affected by the ARPA CTC expansion.

By increasing marginal tax rates among low-income households most likely to experience cliffs, the ARPA CTC expansion increases their prevalence. The current CTC reduces marginal tax rates in the phase-in earnings region, and lowers the cliff rate and gap by 1.0% and 0.6%, almost exactly an offset of the ARPA reform.

Were the US to restore the American Rescue Plan’s expanded Child Tax Credit, two in five Americans would benefit, poverty would fall 9%, and a broad measure of inequality would fall 2%. PolicyEngine also computes impacts for any state; for example, the ARPA CTC would provide $10 billion to Texans (the cost to the US), benefit 46% of Texans, and cut Texas’s child poverty rate by 32%. These impacts, as well as the $100 billion cost and higher marginal tax rates, would likely produce macroeconomic effects that PolicyEngine does not yet model.

With PolicyEngine, you can model highly customizable reforms to the Child Tax Credit and other programs — not only those enacted in the past like the American Rescue Plan Act. Try your own policy model at policyengine.org.

max ghenis
PolicyEngine's Co-founder and CEO

© 2025 PolicyEngine. All rights reserved.