Unbanked, Underbanked, or Something Else Entirely?

In 2009, the FDIC’s first national study of unbanked and underbanked households reshaped how policymakers, researchers, and financial institutions understood financial inclusion.¹ But in the 17 years since, the financial services landscape has changed dramatically. 

Today, digital and mobile banking have made it easier than ever to open and use an account, often without ever visiting a branch. At the same time, nonbank financial services have become embedded in daily life. More than three-quarters of U.S. households use products such as peer-to-peer (P2P) wallets; earned wage access (EWA); and buy now, pay later (BNPL). 

In this new landscape, what does it really mean to be unbanked or underbanked today? 

Drawing on new data from the FinHealth Spend research initiative, this brief examines a variety of ways that definitions of financial inclusion could be revised to account for the evolving product landscape—and why updated metrics are key for advancing financial health.

The size of the underbanked population varies, depending on which products are considered costly or risky.
Share of U.S. households considered underbanked using alternative underbanked definitions.

Notes: FinHealth Spend 2025 survey data. N = 5,096.

This research is made possible through the financial support of Wells Fargo’s Banking Inclusion Initiative. Launched in 2021, Wells Fargo’s Banking Inclusion Initiative provides easier access to low-cost banking and financial education for underbanked and unbanked communities. An estimated 24.6 million U.S. households are underbanked or unbanked, often resorting to higher-cost, less secure alternatives to manage their money. For more information, please visit: https://www.wellsfargo.com/about/inclusion/banking-inclusion-initiative/.

All survey data are drawn from the 2025 FinHealth Spend survey, with data collected between January 8-February 5, 2025, using the Understanding America Study (UAS) Panel. UAS is a probability-based panel of consumers age 18+ designed to be representative of the U.S. population. The study included responses from 5,216 consumers from unique households (margin of error +/- 1.4pp).