COBOLing Together UI Benefits: How Delays in Fiscal Stabilizers Affect Aggregate Consumption

• In Event: Unemployment Insurance, Administrative Capacity, and Labor Market Outcomes

Abstract

The COVID-19 pandemic generated an unprecedented surge in unemployment insurance (UI) claims beginning in March 2020, placing extraordinary strain on state administrative systems. Many state UI programs relied on outdated information technology, including legacy systems written in the programming language COBOL, which limited their ability to process claims efficiently. This paper examines how differences in administrative capacity affected the effectiveness of UI as an automatic stabilizer during the pandemic. Specifically, I compare consumption dynamics across states that relied on antiquated UI systems and those that had modernized their systems prior to 2020.

To measure consumption outcomes, I use high-frequency, state-level credit and debit card transaction data that capture aggregate consumer spending throughout 2020. I combine these data with administrative records on UI claims processing to measure delays in benefit disbursement, including the share of claims delayed by more than 70 days. States are classified as having antiquated UI systems based on documented reliance on COBOL-based infrastructure prior to the pandemic.

I estimate the causal effects of administrative capacity using a two-way fixed effects (TWFE) empirical strategy that compares changes in consumption across states before and after the onset of the pandemic, while controlling for state and time fixed effects. This approach isolates differential consumption responses in states with antiquated UI systems relative to those with modernized systems. The estimates indicate that from March through December 2020, states with antiquated UI systems experienced a 2.8-percentage-point larger decline in card-based consumption relative to states with more modern systems.

To examine mechanisms, I analyze administrative data on UI claims processing and find that states with antiquated systems experienced a 2.1-percentage-point increase in the share of claims delayed by more than 70 days. These results provide direct evidence that outdated administrative infrastructure contributed to delays in benefit receipt. Such delays likely reduced the effectiveness of UI as a fiscal stabilizer by shifting benefit receipt to later periods, when household marginal propensities to consume were lower. As a result, delayed payments generated weaker short-run demand support, contributing to persistent consumption shortfalls in affected states.

Back-of-the-envelope calculations suggest that the failure to modernize UI administrative systems reduced aggregate output by at least $40 billion (2019 dollars) during 2020. These findings highlight the central role of administrative capacity in determining the effectiveness of fiscal policy, particularly during economic crises. More broadly, the results suggest that investments in modernizing public-sector information technology can strengthen the macroeconomic stabilization role of transfer programs and improve the responsiveness of safety-net policies during periods of elevated economic distress.

Author

• Michael Alexander Navarrete, Federal Reserve Bank of Atlanta
  Presenting Author