State-level data on TANF policy changes during the COVID-19 pandemic

Abstract

Objective

COVID-19 mitigation measures prompted many states to revise the administration of their welfare programs. States adopted policies that varied across the U.S. to respond to the difficulties in fulfilling program requirements, as well as increased financial need. This dataset captures the changes made to Temporary Assistance for Needy Families (TANF) programs during the COVID-19 pandemic, from March 2020 through December 2020. The authors created this dataset as part of a larger study that examined the health effects of TANF policy changes during the COVID-19 pandemic.

Data description

TANF is the main cash assistance program for low-income families in the U.S., but benefits are often conditional on work requirements and can be revoked if an individual is deemed noncompliant. Structural factors during the COVID-19 pandemic made meeting these criteria more difficult, so some states relaxed their rules and increased their benefits. This dataset captures 24 types of policies that state TANF programs enacted, which of the states enacted each of them, when the policies went into effect, and when applicable, when the policies ended. These data can be used to study the effects of TANF policy changes on various health and programmatic outcomes.

Objective

Although funded by a federal block grant, TANF is a state-run cash assistance program. States have flexibility on how to run their TANF program, which creates high levels of variability between states in terms of generosity, administrative burden, and punishments for non-compliance [1]. As a block grant program, TANF funding does not change easily in the short-term, especially when there is no change on the federal level. However, during the COVID-19 pandemic, beneficiaries faced additional barriers to receiving assistance due to COVID-19 mitigation measures [2, 3]. For example, TANF assistance is commonly conditional on meeting work activity requirements [4]. Meeting these requirements became more difficult during the COVID-19 pandemic as stay at home orders proliferated, businesses closed, and individuals attempted to keep themselves and their families safe from the virus. As a result, some states modified their TANF practices to avoid beneficiaries losing benefits during a time of great need [2]. This dataset captures the varying policies and timelines of implementation by state. Specifically, the dataset contains information on whether each of the 50 states implemented changes to TANF policy between March 2020 and December 2020.

Researchers can match this dataset with other data on economic and health outcomes to understand how variations in TANF policy can affect economic and health outcomes. Policymakers can use these data to inform the adoption or modification of policies to best serve their constituents.

Dore et al. [5] analyzed this dataset to understand the effect of TANF policy changes during COVID-19 on stress-related health outcomes. They found that overall, more supportive policies led to better stress-related physical, mental, and behavioral health outcomes.

Data description

The dataset is a table that contains the 24 TANF policy categories that represent changes to TANF state policy during the COVID-19 pandemic, the state names and their FIPS codes, and if the states enacted the policies, the start date, and when applicable, the end date. If the cell is empty in the dataset, the state did not implement that particular policy. Examples of policy categories include waiving in person interviews, pausing or lifting existing sanctions, automatically extending or recertifying benefits, allowing participants who reached time limits during the pandemic to continue receiving benefits, and providing additional temporary payments to families. Since the dataset describes the period of March 2020 through December 2020, we created two datasets to represent the ongoing nature of the pandemic. There is only one small difference between the two datasets. One dataset is ready for analysis and includes “December 2020” in the end date cells for policies that had not ended by December 2020. However, these same cells are labeled as “potentially ongoing as December 2020” in the second dataset to clarify the fluid nature of these policies for researchers who may want to extend the analysis beyond December 2020.

To compile this dataset, we first gathered the policy categories from the Center for American Progress website [2]. We then systematically confirmed the implementation of the policy for each of the states listed on that website. The author of the original article, Justin Schweitzer, provided us with a comprehensive list of his sources. We confirmed first using these sources for almost all of the policies. We were unable to confirm some of the policies because his sources came from personal emails or from websites that no longer existed, but we trusted their accuracy. Next, we moved on to finding the start and end dates, which the original article did not contain. We went back to the original sources to identify dates where possible and noted them in the table. When those sources did not contain dates (end dates were especially lacking), we checked other sources online. These sources were: governors’ executive orders, local news articles, legislation, state websites, and email correspondence with state TANF staff. Executive orders and state websites were particularly useful. If we did not find information on the state website (including the TANF website, the governor’s office, or legislative records) or through local news channels directly, we broadened our search using a combination of search terms. The search terms included: TANF, the state-specific TANF program name (e.g. Iowa’s TANF program is called “The Family Investment Program”), a descriptor of the policy (e.g. “work requirement”), COVID-19, coronavirus, pandemic, press release, executive order, legislature, governor, local news, etc. Due to lack of documentation, there were several instances when dates had to be inferred or assumed. The way we did this depended on the source of the information, and we describe our process in more detail in the methodology document in the data depository.

Table 1 Overview of data files and data sets

Limitations

Although we did our best to compile a thorough and accurate timeline and record of policy changes during the first several months of the pandemic, there are some limitations to the dataset. First, sometimes it was not possible to find confirmed start and end dates. In these cases, we made informed estimates using a standardized process based on available information. We include more information on this process in the Methodology document available through Emory Dataverse, but in general, we assumed dates based on dates of other policies implemented in that state, or dates of that same policy implemented in other states. Second, we did not look beyond the list of policies provided by the Center for American Progress for additional policies that may have been implemented. It is possible there were other changes to TANF policy that we missed, though we believe the policy dimensions captured to be the most salient. Lastly, our dataset ends in December 2020 to align with the Center for American Progress list of policies. This period is likely the period with the most changes due to the urgency surrounding the first few months of the pandemic, though we are aware of other policies that were implemented afterwards. For example, more states have since provided emergency cash assistance and the federal government has implemented new TANF policies. However, unemployment was at its highest in the US between March 2020-December 2020 [6] and while TANF caseloads increased beginning in March 2020, they returned to pre-pandemic levels by November 2020 [3]. Thus, we believe our dataset captures the most important time during which individuals needed the most help.