Increasing Take-Up of the Earned Income Tax Credit

Fieldwork by:
Sample:
1,131,122 low-income individuals
Chronologie:
2020
Target group:
  • General
Outcome of interest:
  • Taxation
Intervention type:
  • Nudges and reminders
AEA RCT registration number:
AEARCTR-0005734
Partners:

Administrative burdens limit an individual's ability to access public programs, such as the Earned Income Tax Credit (EITC). Across six randomized evaluations, researchers studied the impact of low-cost, low-touch informational interventions, or “nudges,” on federal and state EITC take-up in California. They found that none of the interventions had an impact on EITC take-up, suggesting that information alone may be insufficient to overcome administrative burdens. 

Policy issue

Even when eligible for government assistance, many low-income families and individuals face barriers to accessing the public safety net. These barriers, or “administrative burdens,” fall into three broad categories: learning costs, psychological costs, and compliance costs.1 Learning costs result from the complexity of accessing and being aware of public programs. Psychological costs result from the stigma around receiving government assistance. Compliance costs result from the effort required to submit paperwork and the time required to navigate the system. 

Administrative burdens can limit the accessibility and impact of government assistance. Moreover, they can exacerbate inequities. Programs that are used more often by high-income and white households are typically much simpler to access than programs that primarily help low-income households, which are more likely to include people of color, people with disabilities, and LGBTQ+ people. Therefore, reducing administrative burdens may increase the accessibility of government assistance and advance equity. 

A wealth of literature on decreasing administrative burdens and increasing public program take-up focuses on “nudges”: low-cost interventions that provide directions, reminders, or other light-touch information with the aim of changing behavior. Nudges have been shown to increase application and enrollment in many public programs.2 However, most nudge-based interventions have focused on people who are already interacting with public programs in some way. Less is known about the impact of nudges on individuals with less government interaction. 

Context of the evaluation

The Earned Income Tax Credit (EITC) lowers the federal taxes that low- and middle-income individuals owe and may give money back after tax filing. EITC distributes more than $60 billion to over 20 million low-income families annually, with an average benefit of $2,500. In California, CalEITC supplements the federal credit. Despite these benefits, it is estimated that one-fifth of all eligible households do not claim the credit, and the take-up rate among EITC-eligible CalFresh  (California’s nutrition assistance program) participants is 50 percent for the CalEITC and 70 percent for the federal EITC.3

Administrative burdens may pose a barrier to claiming EITC. Although EITC is thought to carry less stigma than other programs, there may still be psychological costs attached to mistrust in the government. The filing process can be confusing, complex, and burdensome, leading to compliance costs. In order to claim the credit, households must overcome the learning costs associated with discovering the credit and determining whether they are eligible. For low-income households who are already less likely to file tax returns, these learning costs may be even higher.

Tax forms for the Earned Income Credit.
Photo: Shutterstock.com

Details of the intervention

Researchers conducted six large-scale randomized evaluations during the 2018 and 2019 tax seasons to evaluate the impact of nudges on federal and/or state EITC take-up in California. To better understand if and how administrative burdens contribute to EITC take-up rates, each study sought to address different costs by altering the content and format (i.e., text vs. letter) of nudges sent to study participants. Researchers aimed to address learning costs by providing information on the credit and likely benefit amount, compliance costs by offering application assistance or providing avenues to find assistance, and psychological costs by varying the source of the message (i.e., government vs. non-governmental organization) and the formality of the message. The information was available in languages other than English in four out of the six studies.  

The different messages were delivered by public agencies—the California Franchise Tax Board and the California Department of Social Services—and by a non-governmental organization (NGO) that receives funding from the state to conduct EITC outreach. 

Four studies identified study participants through a private marketing firm, with one study focusing specifically on individuals who had not filed taxes in the past three years. Two studies used CalFresh administrative records to identify study participants. 

To measure subsequent take-up of federal and state EITC, researchers reviewed California tax returns and assessed whether study participants had filed a tax return and if the return included a claim for either the federal or state EITC. 

Table  1 Study Descriptions

Study and sample size

Treatment

Costs addressed and how

Study 1

N=639,244

One treatment arm, delivered as text messages sent by NGO

Learning:
Simple message

Study 2

N=96,370

Four treatment arms, delivered as different letters sent by the state government and NGO

Learning: 
Simple message vs. average benefit amount

Psychological: 
Government vs. NGO messenger

Formal vs. informal

Study 3

N=1,084,018

Four treatment arms, delivered as different text messages sent by NGO

Learning:
Simple message vs. average benefit amount

Compliance:
Web vs. text vs. phone-based assistance

Study 4

N=204,285

Eight treatment arms, delivered as different letters sent by state government

Learning:
Simple message vs. average benefit amount

Compliance:
Local in-person free tax preparation information

Psychological:
Formal vs. informal formatting

Study 5

N=38,093

One treatment arm, delivered as text messages sent by county welfare office

Learning:
Personalized benefit amount

Compliance:
Free tax preparation website

Address of local in-person free tax preparation assistance

Study 6

N=47,104

Three treatment arms, delivered as text messages sent by county welfare office

Learning:
Average vs. personalized benefit amount

Results and policy lessons

Researchers found no evidence that the nudge interventions affected households’ likelihood of filing a tax return or claiming either the federal or state EITC.  

Participants engaged with websites linked in both texts and letters, indicating that individuals did receive and interact with the message. However, the lack of subsequent tax filing may indicate that the hurdles of submitting a tax return, particularly for first-time filers, require more than information and pointers to existing assistance to overcome. 

Further research can seek to understand whether other kinds of messaging, repeated contacts, or other mechanisms for reducing compliance and psychological costs may have more of an impact on populations with limited existing interaction with the government. 

1.
Schweitzer, Justin “How to Address the Administrative Burdens of Accessing the Safety Net,” Center for American Progress, September 14, 2021, https://www.americanprogress.org/article/how-to-address-the-administrative-burdens-of-accessing-the-safety-net/.
2.
Domurat, Richard, Isaac Menashe, and Wesley Yin. “The Role of Behavioral Frictions in Health Insurance Marketplace Enrollment and Risk: Evidence From a Field Experiment.” NBER Working Paper #26153, August 2019. See also: Finkelstein, Amy, and Matthew J. Notowidigdo. 2019. "Take-up and Targeting: Experimental Evidence from SNAP". The Quarterly Journal of Economics 134 (3): 1505–1556.
3.
Iselin, John, Taylor Mackay, and Matthew Unrath. Measuring take-up of the California EITC with state administrative data. No. 2021-3. California Policy Lab Working Paper, 2021.