Introduction

Congress expanded the child tax credit for one year—2021—as part of the American Rescue Plan Act (ARPA; P.L. 117-2).¹ The ARPA expansion of the child credit has often been discussed among its supporters as both a near-universal benefit to help families with the costs of raising children and a program to reduce child poverty.² Prior research has indicated that near-universal direct cash assistance to families with children can have significant impact on reducing child poverty and material hardship in the United States.³ To inform any potential debate about whether and to what extent Congress should extend the ARPA changes to the child credit, this report provides estimates of the potential impact of the ARPA child credit expansion on families' financial well-being prior to the Coronavirus Disease 2019 (COVID-19) pandemic. In other words, the analysis in this report may help policymakers answer the question "what impact could the ARPA-expanded child tax credit have on families' financial well-being and poverty if it were enacted as a standalone provision in a nonrecessionary economy?"

CRS estimates presented in this report suggest the following:

• The ARPA expansion of the child tax credit results in the credit becoming a near-universal benefit generally available to all but the highest-income families with children.
• The ARPA changes to the child credit provide the largest benefits to the lowest-income families, significantly reducing both the prevalence of child poverty (i.e., poverty rates) and the depth of poverty among families with children (i.e., the poverty gap).

This report is structured to first provide an overview of the ARPA changes to the child tax credit for 2021. Next, to analyze the impact of these changes on families' financial well-being, the report provides estimates of the share of families that receive the child credit both before and after the ARPA expansion, followed by estimates of the ARPA expansion's impact on average family income. The report then provides estimates of the ARPA expansion's impact on the prevalence of child poverty (i.e., child poverty rates) and the depth of poverty among families with children (i.e., the poverty gap). To contextualize these impacts, this report also discusses the potential increase in the budgetary cost of the ARPA-expanded child tax credit. Appendix A provides an overview of the methodology and data used in this report. Appendix B and Appendix C provide detailed data cited throughout this report.

The ARPA Expansion of the Child Tax Credit

ARPA made several temporary changes to the child tax credit in effect for 2021 only. Under current law, the credit is scheduled to revert to the parameters in place before ARPA beginning in 2022. (Some of these parameters were enacted as part of P.L. 115-97, commonly referred to as the Tax Cuts and Jobs Act or TCJA, and are scheduled to expire at the end of 2025.)⁴

The Child Credit Before ARPA

For 2021, prior to ARPA, the child tax credit allowed eligible taxpayers to reduce their federal income tax liability by up to $2,000 per qualifying child. A qualifying child was generally any dependent child under 17 years old. The credit was reduced in value, or phased out, by $50 for every $1,000 of income over $200,000 for most unmarried taxpayers or $400,000 for married couples who file joint tax returns (i.e., a 5% phaseout rate).

If a taxpayer owed less in income taxes than the maximum value of the child tax credit, the taxpayer could receive all or part of the difference as the refundable portion of the credit. The refundable portion of the child tax credit—the amount greater than income taxes owed—is often referred to as the additional child tax credit (ACTC) and is calculated using what is sometimes referred to as "the earned income formula." Under the earned income formula, the ACTC gradually increased, or phased in, as earned income rose above a starting threshold of $2,500. Hence, low-income taxpayers with income under $2,500 were not eligible for the credit. The maximum amount of the ACTC was $1,400 per qualifying child. CRS estimates that about one in every five taxpayers (19%) with a credit-eligible child had low incomes that resulted in them receiving less than the maximum credit (i.e., less than $2,000 per child).⁵ After 2021, the ARPA expansion described in this report is scheduled to expire, and the credit is to revert to these "prior law" parameters until the end of 2025.

The ARPA Child Credit Expansion

The American Rescue Plan Act of 2021 (ARPA; P.L. 117-2) makes several temporary changes (for 2021 only) that expand the child tax credit, primarily for low-income taxpayers. These changes include the following:

• Expanding eligibility to 17-year-olds: The law increases the maximum age for an eligible child from 16 to 17 (there are no other changes to eligibility requirements).⁶
• Making the credit fully refundable: The law eliminates the ACTC phase-in based on earned income and eliminates the ACTC cap of $1,400 per child. Hence, the child credit is "fully refundable," and the full value is available to otherwise-eligible taxpayers with no earned income.⁷
• Increasing the credit for low- and moderate-income taxpayers, with larger increases for younger children: The law increases the maximum amount of the credit from $2,000 per child to $3,600 per child for a young child (0-5 years old) and $3,000 per child for an older child (6-17 years old). Generally, this increase in the maximum child credit—of $1,600 per young child and $1,000 per older child—gradually phases out by $50 for every $1,000 of income over specified thresholds until the credit amount equals the current-law maximum of $2,000 per child (i.e., a 5% phaseout rate).⁸ These thresholds (sometimes referred to as the "ARPA thresholds") are
• $75,000 for single filers;
• $112,500 for head of household filers, and
• $150,000 for married joint filers.

The actual income level at which the credit phases down to $2,000 per child depends on the number and age of qualifying children. For many families, the credit plateaus at its pre-ARPA level of $2,000 per child over a range of income, before phasing out when income exceeds the current-law thresholds of $200,000 or $400,000 for married joint filers. These second thresholds are sometimes referred to as the "TCJA thresholds" because they were included in the law commonly referred to as the Tax Cuts and Jobs Act (TCJA; P.L. 115-97). For larger families, the credit may never plateau at the $2,000 per child level over a range of income, but simply continue to gradually phase out.⁹ A stylized example of the child credit benefit amount by income, before and after ARPA, for an unmarried taxpayer with one young child is provided in Figure 1.

Figure 1. Child Credit Amount by Income in 2021 Before and After ARPA Unmarried Taxpayer with One Young Child

Source: CRS calculations based on Internal Revenue Code §24 and P.L. 117-2. Notes: A stylized example assuming the taxpayer has one qualifying child and all income is earned income, with no other sources of income and no above-the-line deductions claimed. Unmarried taxpayers with child credit-qualifying children are assumed in this example to file as head of household. For more examples, see CRS Insight IN11613, The Child Tax Credit: Temporary Expansion for 2021 Under the American Rescue Plan Act of 2021 (ARPA; P.L. 117-2).

As illustrated in Figure 1, an unmarried taxpayer with one young child and income under $144,500 will receive a larger credit from the ARPA expansion, and will receive the full benefit (i.e., the additional $1,600 per young child) when income is under $112,500. Since a majority of taxpayers have low and moderate incomes—95% of head of household filers and 57% of married joint filers had income under $100,000 in 2018 according to data from the Internal Revenue Service (IRS)—a large share of families with children across the income distribution are expected to benefit from the ARPA expansion of the child credit.¹⁰

Impact of ARPA Child Credit Expansion

CRS estimates that in a nonrecessionary economy, the ARPA expansion of the child credit will result in nearly all families with children, including the lowest-income families with children, receiving the child credit. While estimates suggest nearly universal benefit receipt as a result of the ARPA expansion, the estimates also indicate that the largest share of new recipients will be the poorest families. CRS's analysis indicates that the largest increases in income are estimated to occur among poor families with children, substantially reducing the prevalence of child poverty (i.e., child poverty rates) and the depth of poverty among families with children (i.e., the poverty gap).

Family as the Unit of Analysis

Although federal income tax provisions like the child tax credit affect taxpayers (sometimes referred to as "tax units"),¹¹ the impact of the ARPA child tax credit expansion is analyzed in terms of families. This analysis is done at the family level because families can share many resources (like an expanded child credit) and expenses.¹² (Poverty analysis is also generally conducted at the family level.) In some cases, like multigenerational families, a family may be composed of multiple taxpayers. In these cases, any increase in the child credit from the ARPA expansion is aggregated over all taxpayers in the family to determine the ARPA expansion's impact on the family's income. Generally, families discussed in this report are families with children (i.e., they include at least one child under the age of 18).

Family Income Level

Throughout this report, the impact of the child credit is broken down by family income level. In this report, family income level is calculated as the ratio of a family's disposable income to their poverty threshold.¹³ This categorization groups families of different sizes, compositions, and income together by similar standards of living. Families are categorized by their family income level before the ARPA expansion to allow comparison of the same families before and after the ARPA expansion.¹⁴ Notably, this definition of income differs from the definition of income used to phase out the credit (which, for most taxpayers, is their AGI).¹⁵

Change in the Share of Families Receiving the Child Credit

CRS estimates that as a result of the ARPA expansion of the child credit, the share of families with children that receive the credit will increase from 84% to 96% (as illustrated in Figure 2). The largest increase in child credit receipt is estimated to occur among the lowest-income families. Prior to ARPA, about half (52%) of poor families with children were estimated to receive the child credit. After ARPA, the share of poor families receiving the credit is estimated to increase to 94%, a level comparable to benefit receipt among moderate-income families (as illustrated in Figure 2), albeit slightly lower than the overall percentage (96%).

Figure 2. Estimated Share of Families Receiving the Child Tax Credit Before and After the ARPA Expansion of the Child Credit, by Family Income Level

Source: CRS analysis of TRIM3-augmented CPS ASEC files representing data from 2015 to 2017. Notes: See the notes for Table B-3 and Table C-1. The slopes of the lines in the slope graphs reflect the absolute change before and after ARPA, in percentage point terms.

The near-universal receipt of the child credit after the ARPA expansion suggested by these estimates is largely influenced by expansion in credit receipt among poor families. And the substantial increase in estimated credit receipt among the poor families is likely driven primarily by the expansion of credit receipt among the poorest families.

Prior to ARPA, very poor families—those with earned income below the $2,500 earnings threshold—were not eligible for the child credit. By making the credit fully refundable, ARPA effectively eliminated the $2,500 threshold. Hence, these estimates suggest that full refundability is a key factor in expanding credit eligibility to the poorest families.¹⁶ Insofar as recipients with incomes under the $2,500 earnings thresholds are part of families with incomes 100%-199% of poverty (e.g., multigenerational families), the elimination of the earnings threshold could contribute to the modest increase in receipt rates among these families. (The increased eligibility age to include 17-year-olds may also expand eligibility for some poor families.)¹⁷

One reason why poor families may still be less likely to receive this near-universal benefit than moderate-income families (94% compared to 98%-99%, as illustrated in Figure 2) is the ineligibility of certain noncitizen children, specifically noncitizen children without Social Security numbers (SSNs).¹⁸ Under current law, between 2018 and 2025, taxpayers can only receive the credit, including the ARPA-expanded credit, for a child with an SSN. All U.S. citizens, including children, generally receive SSNs. If noncitizen children without SSNs are disproportionately concentrated among poor families, a lower share of poor families would be expected to receive the child credit (both before and after the ARPA expansion) in comparison to moderate-income families.¹⁹

A large proportion of moderate-income and some higher-income families with children are estimated to receive the credit, both before and after ARPA.²⁰ However, the highest-income families (those with income 400% of poverty or more) are least likely to receive the child credit before and after ARPA because of the credit's second or "TCJA phaseout," which was retained by the law. Among moderate- and higher-income families, the expansion of the eligibility age to include 17-year-olds may be a factor in the comparatively modest increases in credit receipt.²¹ Another factor that may drive the increase in credit receipt among some higher-income families is that some poor taxpayers who receive the increased child credit from the ARPA expansion may be part of higher-income families. For example, a low-income couple with a young child may live with higher-income relatives. If the couple were ineligible for the child credit before ARPA, and eligible for the full benefit after ARPA, the family's income would increase by $3,600.

Change in Average Family Income

CRS estimates that the ARPA expansion of the child credit will almost double the average credit amount, from $2,597 to $5,086 (as illustrated in Figure 3). The largest increase is estimated to occur among the lowest-income families. Prior to ARPA, poor families with children were estimated to receive an average credit of $976. After ARPA, these same families are estimated to receive an average credit of $5,421, a credit amount comparable to that received by all but the highest-income families with children (as illustrated in Figure 3).

Figure 3. Estimated Child Credit Amount per Family Before and After the ARPA Expansion of the Child Credit, by Family Income Level

Source: CRS analysis of TRIM3-augmented CPS ASEC files representing data from 2015 to 2017. Notes: See the notes for Table B-4 and Table C-3. The slopes of the lines in the slope graphs reflect the absolute change before and after ARPA, in dollar terms.

Moderate-income and some higher-income families also tend to receive a larger credit after ARPA, although the estimated increase in the credit amount is smaller compared to poor families, as illustrated in Figure 3. For example, on average a family with income between 200% and 299% of poverty is estimated to see their child credit increase from an average of $3,096 per family to $5,288. Even the highest-income families, those with incomes 400% or more of poverty, are estimated to see their credit increase by about $700 per family (from $2,623 to $3,343).

Because these estimated average credit amounts, before and after ARPA, are calculated per family, differences in family composition (i.e., the number of children and their ages) at different income levels could also affect the estimates in Figure 3. CRS's analysis, however, suggests these effects may not be large.²²

Poor families with children tended to receive little to no benefit from the child credit prior to ARPA, either because their low levels of income made them ineligible entirely (as previously discussed) or placed them in the credit's phase-in range (see Figure 1). ARPA substantially increased the benefit amount for poor families by eliminating the phase-in of the child credit (i.e., making the credit fully refundable), as well as creating a larger maximum benefit available to all low- and moderate-income families with qualifying children ($3,600 for young children and $3,000 for older children). For example, prior to ARPA, a single parent with $10,000 of earned income and one young child could receive a $1,125 child credit, as illustrated in Figure 1. After the ARPA credit expansion, they can receive a $3,600 credit, a 220% increase. Generally, the lower the family's income, the larger the increase from the ARPA expansion. For example, if the same single parent with one young child had no earned income, their credit would increase from $0 before ARPA to $3,600 after the ARPA expansion, as shown in Figure 1.

As previously discussed, ARPA also increased the share of poor families receiving the credit (Figure 2). Hence, the larger credit for many poor families (primarily due to full refundability and a larger per-child credit amount) and the larger share of poor families eligible for the credit (primarily due to full refundability) are major factors driving a larger average credit among poor families.

Some moderate-income and higher-income families with children also receive a larger credit on average as a result of ARPA. CRS's analysis suggests that among moderate- and higher-income families, the increases in the credit tend to be smaller on average than for lower-income families. This may be because a significant share of these moderate- and higher-income families are subject to the ARPA phaseout, and thus receive a smaller increase in the credit or no increase at all. For example, as highlighted in Figure 1, some moderate- and higher-income families would receive the maximum credit under ARPA (e.g., $3,600 if income was under $112,500 in Figure 1). Higher up the income scale, families would receive a smaller maximum credit, as the ARPA expansion phases out (e.g., a $2,725 credit for a family with $130,000 of income in Figure 1). Higher-income families are also more likely to include some families subject to the TCJA phaseout (e.g., a $1,500 credit for a family with $210,000 of income in Figure 1), including those who receive no credit (e.g., $0 for a family with $240,000 or more in income in Figure 1), further reducing the average credit per family across the highest-income families.²³

As with low-income families, expanded eligibility for 17-year-olds may increase the credit amount for some moderate- and higher-income families, and hence contribute to a larger average credit per family.²⁴ Some families higher up the income distribution (i.e., with family income greater than or equal to 200% of poverty) may also include low-income taxpayers eligible for the maximum credit amount under ARPA—for example, a low-income couple with a young child who live with higher-income parents or grandparents.

Change in Poverty

Comparing poverty rates before and after the ARPA expansion of the child credit provides one measure of this policy change's antipoverty impact. If the ARPA child credit expansion boosts a poor family's income sufficiently to push them above the poverty threshold (an amount of money below which a family is counted as poor), the family and all its members would be counted as nonpoor as a result of this policy.

In contrast to poverty rates, comparing the aggregate poverty gap before and after the ARPA expansion of the child credit is another metric that can be used to understand the antipoverty impact of this policy change. The poverty gap is the difference between the poverty threshold and a family's disposable income. (By definition, the poverty gap for a family with income above the poverty threshold is $0.) Unlike the poverty rate, which is based on whether a family is above or below the poverty threshold, the poverty gap provides a way of examining the degree to which a family is below that threshold. The larger the poverty gap, the poorer the family.²⁵ For this analysis, poverty gaps are summed together across all poor families to determine the aggregate poverty gap both before and after the ARPA expansion of the child credit.

Change in Poverty Rates

CRS estimates that the ARPA expansion of the child credit will nearly halve the child poverty rate (as illustrated in Figure 4). Similarly, CRS estimates that the ARPA child credit expansion will nearly halve the poverty rate of all individuals (adults and children) who live in families with children (as illustrated with the light pink line in Figure 4).

Figure 4. Estimated Child and Overall Poverty Rates Before and After the ARPA Expansion of the Child Credit by Race/Ethnicity

Source: CRS analysis of TRIM3-augmented CPS ASEC files representing data from 2015 to 2017. Notes: See notes for Table B-6 and Table C-4. The slopes of the lines in the slope graphs reflect the absolute change before and after ARPA, in percentage point terms.

Prior to ARPA, child poverty rates in the United States were higher among Black and Hispanic children compared to the overall child poverty rate (18% and 20%, respectively, compared to 13% of children overall, as illustrated in Figure 4). As a result of the ARPA child credit expansion, the poverty rates of Black, White, and Hispanic children are estimated to fall in roughly similar proportion to each other and in relation to the rate among children of all races and ethnicities (44%, 43%, and 40%, respectively, compared to 46% overall).²⁶ Hence, while the ARPA child credit expansion significantly reduces child poverty rates, it does so roughly proportionally among White, Black, and Hispanic children, and so disparities in poverty rates remain after this policy change. In contrast, the ARPA expansion of the child credit has a comparatively modest effect on the poverty rates of Asian children (a 27% decline), which is estimated to increase the disparity in child poverty rates between Asian children and children overall.

Whether and to what degree the ARPA expansion reduces child poverty rates overall and among children of different races and ethnicities is a result of numerous factors, including the share of poor families eligible for the ARPA-expanded credit, the additional benefit from the ARPA expansion, the depth of poverty families are in before the ARPA expansion, and the child poverty rates before the ARPA expansion.

Previous analysis in this report suggests that almost all poor families with children receive the child credit after the ARPA expansion (94% in Figure 2) and that the ARPA expansion provides an average estimated increase in income of $4,445 among poor families with children (Figure 3). Hence, these estimates suggest most poor families will see a substantial income boost from the ARPA expansion. (Some poor families may not receive the benefit even after the ARPA expansion due to restrictions on the eligibility of certain noncitizen children, specifically noncitizen children without SSNs.)²⁷

The closer a family is to the poverty threshold pre-ARPA, the more likely it is that the ARPA child credit expansion provides the family with sufficient resources to move them above the poverty threshold and out of poverty. In other words, the deeper a family with children is in poverty, all else being equal, the less likely the ARPA expansion will increase their family income enough so that the family (and all its members, including children) exits poverty.²⁸

The depth of poverty compared with the increase in income from the ARPA expansion may be one reason that, even after ARPA, child poverty rates remain high among children of some racial and ethnic groups. For example, prior CRS research suggests that poor Hispanic families and poor Asian families, tend to be deeper in poverty than poor families of other racial and ethnic groups.²⁹ Hispanic and Asian child poverty rates may thus be higher than the overall child poverty rate even after the ARPA child credit expansion because many poor Asian and Hispanic families may require a larger benefit in order to exit poverty. Previous research also indicates that poor Asian families may have fewer children compared to other poor families.³⁰ This would suggest that the ARPA expansion may result in a smaller increase in income per poor Asian family.³¹

High child poverty rates before the ARPA expansion can also be a factor in high child poverty rates after the expansion. For example, among poor Black families with children, CRS estimates suggest a majority receive an increase in the credit after the ARPA expansion (96% see Table C-2), the average increase in the credit received by poor Black families is large ($4,775 see Table C-3), and the depth of poverty is comparable to the all races average.³² But because child poverty rates are much higher for Black children before ARPA, even a comparable reduction in the prevalence of poverty will still mean a larger share of Black children are in poverty compared to all children after the ARPA expansion.

Change in the Poverty Gap

CRS estimates that the ARPA expansion of the child credit will reduce the aggregate poverty gap among families with children by $15.4 billion, or by about 40%, as illustrated in Figure 5. The largest proportional reduction in the poverty gap is estimated to occur among Hispanic families with children (46%) and families of two or more races (43%), with comparable although smaller reductions among Black families (40%) and White families (38%). In comparison, according to CRS's analysis, the ARPA expansion's estimated impact is comparatively smaller among Asian families with children (a 22% reduction from $3.2 billion to $2.5 billion).

Figure 5. Estimated Aggregate Poverty Gap Before and After the ARPA Expansion of the Child Credit by Race/Ethnicity

Source: CRS analysis of TRIM3-augmented CPS ASEC files representing data from 2015 to 2017. Notes: See notes for Table B-7 and Table C-5. The slopes of the lines in the slope graphs reflect the absolute change before and after ARPA, in dollar terms.

As with the ARPA expansion's impact on poverty rates, numerous factors, often in combination, may affect its impact on the aggregate poverty gap. Such factors include the number of families in poverty, the depth of poverty, and the amount of the additional benefit from the ARPA expansion.³³ Differences in these factors may explain some of the differences in the declines in the poverty gap among families of different races and ethnicities. For example, one factor behind the high aggregate poverty gap among White and Hispanic families even after the ARPA expansion may be the large number of poor White and Hispanic families before the ARPA expansion (see Table B-1). Even small poverty gaps per family summed over many families can lead to a large aggregate poverty gap.

The depth of poverty among poor families before they receive the ARPA-expanded credit in comparison to the amount of an increased child credit they receive may also affect the decline in the aggregate poverty gap after the ARPA expansion. For example, as previously discussed, prior CRS research indicates that the typical poor Asian family is significantly poorer than poor families of other races (the typical Hispanic family is also poorer than poor families overall, albeit less so than poor Asian families according to this analysis). Asian families in poverty also tend to receive a smaller credit (perhaps because of fewer or older children or perhaps due to the presence of noncitizen children without SSNs).³⁴ Combined, this may explain why the ARPA expansion is likely to more modestly reduce the aggregate poverty gap among Asian families compared to all families.

Impact on Budgetary Cost

Available data suggest that the ARPA expansion of the child credit will dramatically increase the budgetary cost of this tax benefit. Prior to the ARPA expansion, the total cost of the child credit was $117.7 billion, according to the IRS's most recent data from 2018 tax returns.³⁵ Of that amount, $81.5 billion (70%) offset income taxes owed (the nonrefundable portion), while $36.2 billion (30%) was received as the refundable portion of the credit (and hence exceeded income taxes owed).

The Joint Committee on Taxation (JCT) estimates that the total cost of the one-year ARPA expansion is $105.1 billion, of which $84.4 billion is attributed to the refundable portion of the credit.³⁶ This suggests that the total annual cost of the child credit after the ARPA expansion would increase by almost 90%, potentially to over $200 billion. These estimates also suggest that over $120 billion (60% of the total child credit dollars) could be attributable to the refundable portion of the child credit after the ARPA expansion, a 233% increase in total dollars received as the refundable portion of the credit. By way of comparison, the EITC prior to ARPA was estimated to be about $65 billion in 2018 (changes to the EITC for 2021 will increase its size by about $16 billion).³⁷

As previously discussed, low-income families generally have little to no income tax liability, and receive most if not all of their child credit as the refundable portion. Low-income families include both those in poverty and those with income between 100% and 199% of poverty. CRS estimates of the ARPA child credit expansion's impact suggest that about two-thirds of the increase in aggregate credit dollars will go to low-income families (those with income up to 199% of poverty), with about a third of the increase being received by families with income 200% of poverty or more.³⁸ These estimates suggest that if the ARPA child credit expansion were extended, it would likely be, in budgetary terms, the largest cash assistance program for low-income families with children.

Conclusion

The analysis presented in this report suggests the ARPA-expanded child credit is likely to significantly increase incomes of low-income families with children and substantially reduce both the prevalence of child poverty and the depth of poverty among families with children. Research suggests that a child benefit similar to the ARPA-expanded credit could have significant long-term benefits for children's health, education, and future earnings.³⁹

These effects could be diminished by other factors that are beyond the scope of this report. For example, some evidence suggests that the ARPA-expanded child credit, which working and nonworking parents alike can receive, may discourage low-income parents from working. Existing research suggests this impact may be small, but reductions in earned income could reduce the benefit's anti-poverty impact.⁴⁰ Some policymakers may be concerned with the administration of the ARPA-expanded child credit. Families with children who have never been required to file a federal income tax return due to their low incomes would need to file a return to receive the credit. Failure to do so would mean that these families would not receive the child credit, which could lessen its antipoverty effectiveness. More broadly, the administration of this expanded credit generally requires that half of the credit be issued monthly during the last six months of 2021. To determine how much to advance, the IRS will be relying on older data—from 2020 or 2019—to estimate eligibility and then issue advance payments. If this information is out of date and not corrected, a taxpayer could receive a benefit they are ineligible for, and in some cases, have to pay it back. This could discourage take-up of the advance payment or result in IRS compliance measures, like audits, that may discourage participation.

Even with these limitations, CRS analysis suggests that the ARPA expansion of the child credit will transform the credit into a near universally available benefit that provides significant financial assistance to many low-income families with children. Many of the poorest families—who received little to no benefit from the credit prior to ARPA—will receive the largest increases from this legislative change, resulting in a substantial reduction in child poverty. For families with the highest levels of financial hardship—those deepest in poverty—the ARPA-expanded credit may reduce the depth of their poverty, but be insufficient on its own to lift them out of poverty.

Appendix A. Methodology and Data

CRS used Version 3 of the Transfer Income Model (TRIM) to calculate the estimates presented in this report. This model applies government program rules onto data from the Census Bureau's Current Population Survey Annual Social and Economic Supplement (ASEC), allowing CRS to simulate how changes in government programs may impact the population. For this report, CRS simulated ARPA's child credit changes in TRIM's Federal Income Tax module, and then examined how these changes affected families using definitions of income and poverty associated with the Supplemental Poverty Measure (SPM). For more information on the standards CRS adheres to in conducting TRIM analysis, the ASEC, and the SPM poverty measure, please see CRS Report R46824, Need-Tested Benefits: Technical Companion to Selected CRS Reports on Need-Tested Benefits Receipt by Families and Individuals.

Several other methods and caveats specific to this report are discussed further below.

Analytical Methodology

Some racial and ethnic groups are smaller in population size (and ASEC sample size) than others. Because CRS divided each racial and ethnic group into several income levels for this report, the sample size for some estimates was too small to draw reliable conclusions about using only one year of data. To reduce error introduced by small sample sizes, CRS chose to average three years of data (the 2016-2018 ASEC surveys, which report income and other data from 2015-2017) to produce these estimates.⁴¹ CRS also conducted standard error analysis⁴² on these three-year averages to ensure that the estimates were sufficiently precise. Although averages tend to skew towards extreme values (unlike some other measurements of central tendencies, such as medians), CRS chose to calculate averages in order to conduct standard error analysis consistent with Census Bureau guidance.⁴³

Racial and Ethnic Category Methodology

CRS uses responses recorded on the ASEC survey to define racial and ethnic groups. The ASEC asks respondents to self-identify as American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, or White.⁴⁴ Additionally, respondents may identify as more than one race or as some other race. Hispanic ethnicity is addressed separately in another question. Therefore, persons of any racial identity can also identify as being Hispanic.⁴⁵

CRS combines these two survey questions to categorize individuals according to both their race and Hispanic ethnicity. CRS categorizes Hispanic individuals as being Hispanic, regardless of their racial identity. All non-Hispanic individuals are then categorized according to their respective racial identities. Using this definition, CRS further defines the race or ethnicity of a family by the racial and ethnic composition of the family members. If one member reports a racial or ethnic identity that differs from their other family members, the family is categorized as having two or more races. For example, if one member of the family identifies as Hispanic and Black, and the other two members identify as non-Hispanic Black, the family would be considered of two or more races. If all members of the family share the same racial or ethnic identity, they are categorized as that identity.

By using compositional measures to define families' racial and ethnic group, CRS can evaluate families of multiple racial and ethnic identities. For example, a family comprised entirely of Black individuals, none of whom identify as Hispanic, would be considered Black. A family comprised of both non-Hispanic Black individuals and Hispanic Black individuals would be considered two or more races. A family comprised entirely of Hispanic Black individuals would be considered Hispanic. This method of defining an individual's racial and ethnic group preferences Hispanic ethnicity over racial identity, which may not reflect how a respondent views their own racial and ethnic identity.

Exclusion of Small Population Groups from this Analysis

Despite improving overall sample sizes by averaging three years of data, this report does not present results for persons who identify as non-Hispanic American Indian or Alaska Native, or as non-Hispanic Native Hawaiian or Other Pacific Islander, due to their relatively small sample sizes and relatively high standard errors. However, these persons were counted toward the aggregate totals for all analyses present in the report. Additionally, if such individuals lived in a family with two or more racial or ethnic identities present, they were included in that family.

The Family as the Unit of Analysis

Although the ARPA expansion of the child credit affects taxpayers, the impact of these provisions is analyzed in terms of families. A taxpayer is generally composed of all individuals listed on a federal income tax return (IRS Form 1040) and includes an individual, their spouse (if married), and any dependents. In contrast, poverty analysis is done at the family level because families can share many resources and expenses. Hence, in this report analyses of the impact of the ARPA-expanded child credit are generally done at the family level. In this report, a family is composed of people living together related by blood or marriage (the family), cohabiting partners, and foster children. In some cases, like multigenerational families, a family can be composed of multiple taxpayers. In these cases, tax liabilities and/or benefits for all taxpayers are aggregated to determine the impact of the income tax on the family's resources. If a family is determined to be poor, all members of that family are counted as poor.

Family Income Level

This report uses disposable income to categorize the population across separate income levels and compare higher-income families to families living in poverty. Disposable income is defined as the resources available to families under the research Supplemental Poverty Measure (SPM).⁴⁶ These resources are calculated as money income (i.e., earned income plus interest and dividends plus cash social insurance benefits [e.g., Social Security, unemployment insurance]); minus taxes, minus work expenditures (e.g., child care), and minus medical expenditures; plus the value of tax credits (including the child credit and the earned income credit [EITC]) and the value of in-kind benefits (such as food and housing subsidies).

Families with the same disposable income may have differing poverty status (under the SPM, there are thousands of different poverty thresholds which depend on a family's composition, geographic location, and housing tenure).⁴⁷ To account for this and compare families living in poverty to families with higher income levels, CRS computes the ratio of each family's SPM resources to their respective SPM poverty threshold. From this ratio, CRS created five separate family income levels, delineating those with disposable income 0%-99% of their poverty threshold (i.e., families living in poverty) from families with ratios of 100%-199%, 200%-299%, 300%-399%, or 400% or higher.

Child credit amounts are estimated per taxpayer, and then included in family resources.⁴⁸ For example, assume a family includes two taxpayers (e.g., a single parent with her child lives with her sibling and their family). For each taxpayer, TRIM3 estimates their child credit amount (using AGI, if the family is subject to either the ARPA or TCJA phaseout). According to TRIM3 estimates, one taxpayer receives a $500 increase in their child credit from ARPA and the other taxpayer receives a $100 increase in their child credit from the ARPA expansion. The family's resources will thus increase by $600. Because eligibility for the child credit does not affect eligibility for other need-tested programs,⁴⁹ or tax liabilities (since tax credits are not taxable income), the change in a family's resources before and after ARPA will be due entirely to the ARPA changes to the child credit.

Limitations

There are several limitations that should be considered when interpreting the results presented in this report. First, because of the data lags in the TRIM3 model and the use of three-year averages in this analysis, the underlying data used in this report (which reflect a nonrecessionary economic period) are not analogous to economic conditions in 2021, nor will they be for the eventual postrecessionary economy.

Second, the results do not account for the fact that some eligible families may not file an income tax return, which is necessary to receive the child tax credit. Instead, the model used by CRS in this analysis assumes that 100% of eligible families will receive the full value of the credit for which they are eligible.⁵⁰ This limitation may be particularly important for the ARPA child tax credit expansion, which made families with very low or no earned income eligible for the child tax credit. These families are generally not required to file an income tax return, and as a result may not know that they are required to file a return in 2021 to claim the ARPA-expanded child tax credit.

Third, the results do not include estimates of eligibility in U.S. territories. The model used by CRS in this analysis does not include information on residents of U.S. territories. As a result, CRS is not able to estimate eligibility for the child tax credit in the territories.

Fourth, the results do not attempt to estimate the impact of the requirement in ARPA that the Treasury issue half of the expected 2021 child tax credit in periodic payments to families beginning in July 2021. Although the use of periodic payments may change the timing of when families receive benefits from the child tax credit, this change will not affect the annual amount of the credit for which families are eligible.

Fifth, the categorization of racial and ethnic groups may not reflect the way that persons reporting their race and ethnicity self-identify. A Pew study from 2014 found that 67% of persons who identified as Hispanic viewed being Hispanic as part or all of their racial background.⁵¹ The authors noted that this has two implications: (1) that race among Hispanic persons does not fit neatly into the Census Bureau's current racial categorization standard and (2) that many Hispanic individuals actually view themselves as mixed race, and making this distinction would increase the proportion of mixed race persons in the United States by 2 percentage points. However, the complexities of Hispanic identity (or other racial identities) are difficult to understand using the ASEC, due to sample size and survey issues. The method employed in this report attempts to capture those persons who identify as racially Hispanic, and therefore likely incorrectly categorizes some persons whose Hispanic ethnicity does not play a role in their racial identity.

Finally, estimates prepared using survey data (which include these results) tend to underestimate the value of refundable tax credits such as the child tax credit. Research suggests that one cause of this underestimation may be that families with children where multiple adults file an income tax return have an incentive to strategically choose which adult will claim the children on their return.⁵² By doing so, families may maximize the amount of the child tax credit they receive in a way that is not fully captured by the model used by CRS in this analysis.

Appendix B. Pre-ARPA Baseline Estimates

Number and Average Income of Families with Children

Table B-1 presents estimates of the number of individuals in families with children, by family income levels and race and ethnicity. For example, these data indicate there are 264,000 Asian families with children who are living in poverty. These estimates present baseline population counts, before the ARPA child credit expansion.

Table B-2 presents estimates of average family income among families with children, by family income levels and race and ethnicity. For example, these data indicate that among all Asian families with children who are living in poverty, the average family income is $17,427, as defined by their SPM family income. This table presents baseline income before the ARPA child credit expansion.

Grouping families by their ratio of family income to SPM poverty threshold (i.e., "family income level") allows families of similar standards of living across the United States to be grouped together. Families are each characterized by their family income in relation to their appropriate SPM poverty threshold. A family's SPM threshold depends on numerous factors, including where they live (i.e., the geographic variability of housing costs), whether a family rents or owns, and the family's size.⁵³ So, for example, the SPM threshold for a family in an urban area with high housing costs would generally be higher than the SPM threshold for a family living in a rural area with lower housing costs, all else being equal. This could result in a poor family in an urban area with high housing costs having a higher family income than a poor family in a rural area with lower housing costs. Insofar as there are significant differences in poverty thresholds by race or ethnicity—a topic beyond the scope of this report—there hence may be deviations in average family income by race and ethnicity for a given income category.

Families are categorized by their family income level before the ARPA expansion (i.e., the pre-ARPA child credit is included in income, but the increase from the ARPA expansion is not included in income for the purposes of categorizing families by family income level). This categorization allows for a comparison of the same families before and after the ARPA expansion of the child tax credit.⁵⁴

Families with Children Receiving Pre-ARPA Child Credit

Table B-3 presents estimates of the percentage of families with children, of a given income level and race and ethnicity, who received the pre-ARPA child tax credit. For example, these data indicate an estimated 57% of all Asian families with children in poverty (family income level <100% of poverty) received the pre-ARPA child tax credit. Table C-1 displays the average post-ARPA credit across the same income levels and race and ethnicity categories.

Table B-4 presents estimates of the average pre-ARPA child tax credit received by families with children, by income level and race and ethnicity. For example, these data indicate Asian families with children in poverty received a pre-ARPA child tax credit of $1,167 on average. Table C-3 displays the average post-ARPA credit across the same income levels, and race and ethnicity categories.

Poverty

Table B-5 presents estimates of the number of individuals living in poverty in families with children, by work status and individual race and ethnicity. Additionally, the table displays the number of children living in poverty separately. For example, these data indicate there are 1.06 million Asian individuals (adults and children alike) living in families with children who are living in poverty. Of these individuals, 428,000 are children. Additionally, roughly 77% of these individuals (816,000 of 1.06 million) live in families with workers. These estimates present baseline poverty counts, before the ARPA child credit expansion.

Table B-6 presents estimates of the percentage of individuals living in poverty in families with children, by work status and individual race and ethnicity. Additionally, the table displays the percentage of children living in poverty separately. For example, these data indicate 11% of Asian individuals (adults and children alike) living in families with children are in poverty. The percentage of Asian children living in poverty is also 11%. These estimates present baseline poverty rates, before the ARPA child credit expansion. Table C-4 illustrates the poverty rate after the ARPA child credit expansion, across the same family categories and race and ethnicity categories.

Table B-7 presents estimates of the poverty gap among families living in poverty with children, by work status and individual race and ethnicity. For example, these data indicate Asian families living in poverty with children would require $3.2 billion on aggregate to exit poverty. These estimates present baseline poverty gaps, before the ARPA child credit expansion. Table C-5 illustrates the poverty gap after the ARPA child credit expansion, across the same family categories and race and ethnicity categories.

Appendix C. Post-ARPA Expanded Child Credit Estimates

Table C-1 presents estimates of the percentage of families with children, of a given income level and race and ethnicity, who received the post-ARPA child tax credit. For example, these data indicate an estimated 84% of all Asian families with children in poverty (family income level <100% of poverty) received the post-ARPA child tax credit. Table B-3 displays average pre-ARPA credit receipt across the same income levels and race and ethnicity categories.

Families with Children Receiving a Larger Child Credit Under ARPA

Table C-2 presents estimates of the percentage of families with children, of a given income level and race and ethnicity, who received a larger child tax credit due to ARPA. For example, these data indicate an estimated 84% of all Asian families with children in poverty (family income level <100% of poverty) received a larger child tax credit due to ARPA, defined as an increase of $10 or more. Table B-3 displays the average pre-ARPA credit receipt across the same income levels and race and ethnicity categories.

Income

Table C-3 presents estimates of the average percentage and dollar increase in income due to ARPA among families with children, categorized by a given pre-ARPA income level and race and ethnicity. For example, Asian families with children in pre-ARPA poverty (family income level <100%) received an average increase of $3,155, with an average percentage increase of 81%.

The percentage increase in income in Table C-3 may not equal the average increase divided by the pre-ARPA income as reported in Table B-2. For example, the percentage increase in income for poor Asian families in Table C-3 is 81%, which does not equal the average dollar increase ($3,155) divided by the pre-ARPA income of Asian families living in poverty as reported in Table B-2 ($17,427). The reason for this disparity between average percentage increase and average dollar increase is that percentage change, in this case, is more susceptible to skew. Prior to ARPA, the lowest-income families did not qualify for the child credit, or qualified for a reduced credit, because of their low incomes. ARPA does not restrict families with low or no earnings from receiving a full credit. For such families, the ARPA credit will represent a larger increase than for families already receiving the full ACTC. Because of their low initial incomes, the percentage increase reported could be relatively large, significantly skewing the average.⁵⁵

Poverty

Table C-4 presents estimates of the percentage of individuals living in poverty in families with children after the ARPA child credit expansion, by work status and individual race and ethnicity. Additionally, the table displays the percentage of children living in poverty separately.⁵⁶ For example, these data indicate 8% of Asian individuals (adults and children alike) living in families with children are living in poverty. The percentage of Asian children living in poverty is also 8%. Table B-6 illustrates the poverty rate before the ARPA child credit expansion, across the same family categories and race and ethnicity categories.

Table C-5 presents estimates of the poverty gap among families living in poverty with children after the ARPA child credit expansion, by work status and individual race and ethnicity. For example, these data indicate that after ARPA, Asian families living in poverty with children would require $2.5 billion on aggregate to exit poverty. Table B-7 illustrates the poverty gap before the ARPA child credit expansion, across the same family categories and race and ethnicity categories.