Disentangling in-work and child tax credits to address improper payments: An incremental approach to comprehensive reform

In 2023, federal agencies collectively identified $236 billion worth of improper payments across all the federal programs with annual spending greater than $10 million. Improper payments include both those that should not have been made because the recipient was ineligible, as well as payments made in the incorrect amount, whether recipients were paid too much or too little. That said, almost three-quarters of improper payments were overpayments in 2023.¹ 

The IRS has identified refundable tax credits as particularly susceptible to improper payments. The Earned Income Tax Credit (EITC) has been plagued by an alarmingly high improper payment rate for decades. The most recent estimates put the scope of improper EITC payments at one-third of all payments, amounting to $21.9 billion in 2023 — more than the federal government spends each year on Temporary Assistance for Needy Families (TANF).² These high improper payment rates have persisted despite years of effort to address them on the part of both the IRS, through administrative reforms, and Congress, through small legislative tweaks.³

The EITC, along with the Child Tax Credit (CTC), are foundational policies for supporting American families. Persistent program integrity issues undermine this vital goal and should be addressed in a manner that protects these tax credits and the families served by them. In this report, we identify the source of the problem, consider proposed solutions from tax experts, and offer incremental recommendations for reducing improper payments in a manner that continues to support families.

There is a broad consensus that the EITC’s improper-payment issues stem from the credit’s complex income and qualifying-child eligibility requirements and the IRS’ limited ability to verify eligibility in a timely manner.⁴ Income misreporting and qualifying-child errors account for the majority of improper payments. While income misreporting is the most common error, qualifying-child errors account for the largest dollar amount of overpayments.⁵ For that reason, we focus on the source of qualifying-child errors here.⁶ 

Both the EITC and CTC allow filers to claim qualifying children if they meet certain eligibility criteria. Different family provisions in the tax code — such as the EITC and CTC, but also the Child and Dependent Tax Credit, Dependent Exemption, and Head of Household Filing Status — have used inconsistent definitions of “qualifying children” in the past. Congress finally addressed this with the Working Families Tax Relief Act of 2014, which adopted a uniform definition of qualifying children, and with the Tax Cuts and Jobs Act, which eliminated the Dependent Exemption and replaced it with a larger CTC.⁷ Standardizing definitions was a good first step, but confusion remains over several aspects of the qualifying-child tests. There are several criteria that children must meet in order to qualify for the credits. Two in particular create complexity that gives rise to improper payments.⁸

Relationship test. The qualifying child must be related to the claimant in one of the following ways: son/daughter, adopted child, qualifying foster child, half-sibling, step-sibling, nephew/niece, or grandchild.

Residency test: The qualifying child must have lived with the claimant for more than half of the tax year. The EITC has an additional requirement that the claimant and child must have resided in the United States for more than half of the tax year.

In the abstract, both are reasonable tests to ensure that the benefits of these tax credits go to the claimants who are caring for the child. The trouble arises when these rules meet the reality of changing family circumstances over time. Households with stable marriages rarely encounter complexity in claiming tax benefits, as there are typically no changes in relationship or residency, and no conflicts between claimants. However, the increasing complexity of family arrangements—single parents with shared custody, married couples with non-biological parents sharing custody, and intergenerational households—makes determining claimant eligibility increasingly difficult.⁹ 

In many cases, the fluid nature of children’s residence over the tax year leaves parents confused as to whether they have met the test for the various family provisions in the tax code. Unmarried parents themselves often disagree about the extent to which each of them lived with their child over the year.¹⁰ Parents sometimes find their claims rejected upon filing because the other parent has already claimed the child. In other cases, parents may shift qualifying-child claims, either because they are not aware of the specific rules, or in order to purposely maximize total credit amounts. For example, the IRS has established tiebreaker rules for situations where the child resides with both parents equally, requiring that the parent with the higher earnings claim the child for EITC purposes. However, because residency is hard to verify, this provides the opportunity to shift the child’s residence on paper to maximize the EITC.¹¹ 

The EITC has substantial churn from year to year. About one-third of EITC claimants in a given year did not claim the EITC the prior year, leaving more of them plausibly unaware of the rules.¹² In contrast, the CTC, which is not targeted to a narrow slice of tax filers, is more likely to be claimed and understood from year to year. 

The problems associated with determining who can claim a qualifying child are exacerbated by the limited ability of parents to document the necessary information and the limited ability of the IRS to adequately verify it. Whereas other family benefits, like social assistance or Medicaid, rely on application processes that verify eligibility before benefits are paid out, family benefits in the tax code, like the EITC and CTC, rely more on a system of self-attestation with limited capacity for third-party verification and a heavy reliance on audits. Qualifying-child errors are particularly tough to monitor because they are “factually intensive and not easily verifiable.”¹³ Audits are often burdensome for tax filers and the IRS alike, and have raised concerns about unfair treatment in recent years.¹⁴ 

As the IRS Taxpayer advocate has noted in congressional testimony, filing for the EITC is essentially an application for the EITC, with verification backloaded rather than upfront as with other family benefit programs. This substantially lowers the administrative cost for the government and for filers, but increases the volume of applications that need to be reviewed for error and fraud. The EITC’s administration costs comprise less than 1 percent of the total benefits delivered, while for other non-tax benefit programs, determining eligibility can comprise as much as 20 percent of the total program cost.¹⁵

The persistence of improper payments related to qualifying-child errors and the absence of efficient mechanisms for verifying eligibility suggest the limitation of administrative reforms to address them. Several government watchdogs, including the Taxpayer Advocate and Government Accountability Office, have pointed to the potential merits of a comprehensive legislative overhaul of the EITC and CTC as a solution.¹⁶

One novel legislative reform, which has been recommended by the Taxpayer Advocate for over a decade — most recently in the 2025 “purple book” — is to restructure the EITC and CTC into two totally distinct credits. In this scenario, the EITC would be stripped down into a straightforward tax credit aimed at boosting the incomes of low-wage workers, irrespective of whether they are caring for children in their household. The children’s portion of the EITC would be shifted over to the CTC in line with its existing structure, aimed at helping families more generally with the cost of raising children. The program-integrity benefits of eliminating the need to determine eligibility for qualifying children for the EITC are straightforward. According to the Taxpayer Advocate:

This change could …substantially reduce improper payments. The IRS receives Forms W-2 and other information reporting documents directly from employers and other payors of income. For that reason, it can accurately verify income amounts for EITC recipients who are employees, who constitute by far the largest group of EITC claimants.¹⁷

Right-sizing the EITC would also help reduce credit-maximizing behavior. By restoring the EITC to its original intended goal of offsetting payroll-tax liability, Congress could ensure that the maximum EITC benefit does not exceed a claimant’s combined income- and payroll-tax liability. This would retain its work incentive effects while minimizing incentives to misreport or misrepresent earnings.

While this would drastically reduce the sources of improper EITC payments, the IRS would still need to verify qualifying children’s eligibility for the CTC. The IRS is already equipped to verify relationships in most cases. One joint IRS-Treasury study found that the IRS could verify the relationship between the claimant and child in 80% of returns using their Social Security numbers. Compliance audits suggest residence tests are a major source of qualifying-child errors.¹⁸ The same study found that the IRS could use data from other government agencies that pre-verify benefits to verify residence in many cases as well.¹⁹ Still, verifying residence for parents not receiving these other benefits may be harder without being intrusive, particularly where there is conflict between claimants. If we think of the residence test as just one way to determine who is the primary caregiver for children, though, we can look to other countries for guidance on how to verify eligibility for claiming tax credits.

Rather than have parents self-attest each year, many other countries allow parents to designate a primary caregiver for the sake of claiming child tax credits and verify that status when the child is born. This includes Australia, Canada, New Zealand, and the United Kingdom.²⁰ Table 1 provides the relevant information for these four countries plus the United States. When parents fill out forms for the country’s vital statistics and social security agencies, this information is shared with the country’s revenue agency, which often administers family benefits as well. These agencies retain this information and can administer credits accordingly. They also have portals where the primary caregiver is required to update the agency when there are any changes in family arrangements (e.g., custody changes). Parents must provide relevant documentation for these changes so that the revenue agency can verify it before adjusting credit payments. This mechanism has the added benefit of reducing barriers to splitting credits depending on custody arrangements.

Each of these countries have modernized their revenue agencies to administer child benefits in a manner that minimizes the need for parents to prove, and agencies to verify, eligibility, leaving the U.S. as the only country that requires parents and the IRS to do both on an annual basis, even if there are no changes in family arrangements. 

Ideally, Congress would simultaneously convert the EITC and CTC into distinct worker and child credits and modernize administration to allow continuous CTC eligibility with documentation, reducing improper payments among refundable tax credit programs. There have been several proposals and bills in recent years along these lines.²¹ However, there is little congressional appetite at the moment for comprehensive overhauls of this nature. While complete distinction remains our north star, the goal may be better served by an incremental approach.

Gradually separating the child and work components of the EITC is both politically practical. Congress has gradually expanded the EITC in the 50 years since its initial passage. At its launch, the EITC was much smaller and targeted solely at families with children. The goal was to offset rising payroll taxes and reduce barriers for parents moving from welfare to work.²² In 1993, Congress expanded the EITC to low-income workers without children and adjusted the family portion to vary with the number of children in the household. Finally, in 2009, the EITC underwent a further refinement to offer three tiers of family benefits for households based on whether they had one, two, or three or more children.

The incremental refinement of the EITC allowed the program to build on previous successes and to adjust benefits in response to the composition of families and the strength of the economy. However, the introduction of the CTC in 1997 and its expansion over time has made the children’s portion of the EITC redundant. Congress took the first steps toward consolidating overlapping family tax benefits in 2017, when the Dependent Exemption was eliminated and replaced with a more generous CTC to keep families whole. It was a successful strategy. 

Refining and disentangling the child portion of the EITC over time makes it easier to prevent families from experiencing sudden swings in their benefits. Our proposal fine-tunes EITC rates to avoid having some families be substantially disadvantaged from one tax year to the next. Additionally, it reduces what would likely be very high immediate costs associated with separating the credits altogether in one reform. Just as Congress had little appetite in 1975 to expand the EITC into what it is today, it is reasonable to assume that Congress might be more apt to take incremental steps toward full separation over time as budget circumstances allow for it. 

As that happens, Congress can work on establishing a new mechanism for parents to claim and the IRS to verify eligibility for the CTC. We propose that IRS modernization follow best practices for administering child tax credits from other countries.²³ The establishment of the IRS Portal during the pandemic was a good first step, but more comprehensive work is required.²⁴ Shifting to a primary-caregiver standard for claiming the CTC is the next logical step.

The process would begin at birth. New parents, when filling out forms at the hospital for the child’s birth certificate and Social Security number, could apply to have the relevant information sent to the IRS to keep on record. Parents could choose who they would like to designate as the primary caregiver at this time so that they would not have to redesignate one each year when they file their taxes. If there were any changes to care arrangements, parents could make updates through the IRS portal with the relevant documentation (e.g., court documents related to foster, adoption, or other custody arrangements, etc.). 

This will reduce administrative burdens on parents as well as the IRS by limiting the need for verification at the start of a child’s life and only updating information when there are relevant changes that require it. Furthermore, the IRS could cross-check data with other programs (e.g., SNAP or Medicaid) to better detect improper payments without the need for traditional audits.²⁵

These additional verification mechanisms would ensure that the problems we see now with the EITC would not simply be shifted over to the CTC as it becomes the main vehicle for supporting families with children. 

There are a variety of ways to shift the child component of the EITC to the CTC, depending on one’s goals. Our incremental approach starts with the goal of transferring the 2009 addition of the third child component of the EITC to the CTC: providing the same value to families through the CTC while minimizing the overall costs of the reform. This effectively provides us with room to shift about $1,000 per child of the EITC over to the CTC.²⁶ We make the following changes to the EITC and the CTC’s maximum amount, thresholds, and phase-in and -out rates (see Tables 2-5 for summaries).

We completely eliminate the EITC category for three or more children and reduce the maximum EITC amount accordingly for workers with one or 2+ children. This allows us to reduce the phase-in rates for the EITC from 34% to 23% for one child and from 40% and 45% for two and three-plus children to 28% for two-plus children. We make minor tweaks to the phase-out rates, adjusting them from 15.98% and 21.06% to a flat 16% and 20%, respectively.

To ensure reforms do not leave families with children worse off, we make corresponding changes to the CTC, increasing the maximum amount from $2,000 per child to $3,000 per child. We collapse the distinction between the refundable and nonrefundable portion of the CTC, making the entire credit refundable. To make up for the loss of some portion of the EITC for low-income families, we begin phasing the CTC in at the first dollar of earnings at an enhanced per-child rate. The flat 15% per-household rate is doubled to 30% per household.  

The result is that a much smaller portion of total tax credits for families with children comes from the EITC and a much larger portion comes from the CTC. Families with children under 18 are held harmless or made better off with these changes.

We used Policy Engine, a microsimulation model of the United States tax and benefit system, to assess the cost and distributional impact of our proposed reforms. Modelling these changes demonstrates that the incremental approach allows almost all families to be better off. It increases net income for about 40 percent of households and decreases it for 2 percent. The net cost of the reforms would be $46.3 billion in the first year.²⁷ Figure 2 breaks down the relative net changes by household income decile.

As expected, most of the relative gains are concentrated among middle and upper income households. These households were not receiving any benefit from the EITC but would see their CTC benefit increase under the reforms. A small number of households see their income decrease, though almost all by less than 5 percent. The incremental approach allows families, for the most part, to be kept whole by the reform, or to have any losses kept to a minimum. Figure 3 looks at the same deciles using absolute changes. The average household sees their income rise by $346. The gains are concentrated among middle and upper income households receiving a larger CTC.

Some of the additional cost associated with the reforms would be offset by savings that potentially flow from a reduction in improper payments. Congress would still need to find additional sources of revenue to offset the full cost of the reform, though.

This year marks the 50th anniversary of the EITC’s introduction. As we reflect on its history, we can acknowledge two truths about the program. It is one of our most beloved programs for supplementing the wages of low-income workers, garnering bipartisan support for most of its history. It is also one of our most inefficient programs for low-wage workers, plagued by decades of high improper payment rates.

We follow a long line of tax experts who identify the source of the problem as verifying qualifying children and recommend shifting to distinct refundable tax credits for workers and for children. Congress has struggled to act on these recommendations, though. Congressional Democrats have shown little appetite for EITC reform relative to a maximalist approach that focuses on expanding both tax credits. In contrast, congressional Republicans have demonstrated their interest in similar reforms with the successful consolidation of the Dependent Exemption and CTC as part of the Tax Cuts and Jobs Act (TCJA) of 2017. This interest and the impending expiration of major portions of TCJA in 2025 offer an opening for making progress on EITC reform.

Comprehensive EITC is the long-run goal but Congress can take the initial incremental steps now with the proposed reforms we lay out here.