How many financial institutions are in the U.S.?

It might seem like a simple thing to establish how many financial institutions are in the United States. After all, this is pretty important when you’re establishing coverage for as many consumers as possible. But over the past several years building Plaid, we’ve learned that this is actually a bit trickier in practice — mostly because M&A activity and institution failures have changed the landscape dramatically in the past several decades, especially post-recession, and because financial infrastructure is itself changing all the time, which sometimes makes it seem like there are more integrations than institutions.

According to the most recent data from the FDIC and NCUA, though—which we think is the most reliable information—there were 5,801 FDIC-insured institutions and another 5,733 NCUA-insured credit unions nationwide. That’s 11,652 total. Yet it’s not uncommon to hear that there are nearly 18,000 financial institutions in the United States.

So, how’d we get here?

To map the history of the number of financial institutions is to trace a long and complicated tale of consolidation and failure across the U.S. banking system. Both banks as well as credit unions are currently at their lowest numbers since regulators started tracking these metrics. FDIC-insured financial institutions reached their peak of 17,900 in 1984, while the number of credit unions apexed in 1969 at 23,866.

Since then, the landscape has been disproportionately reshaped by the exits of smaller institutions. From 1984 to 2011, more than than 10,000 banks left the industry during that period as a result of mergers, consolidations or failures (17 percent of the banks collapsed). From 1970 to 2011, there were an estimated 13,000 credit union mergers, according to the Federal Reserve.

Today, there tends to be even more concentration at the top: For example, the top 200 banks and credit unions comprise about 70 percent of all U.S. depository accounts. But small banks and credit unions remain critical for rural areas and other constituencies, such as teacher’s unions, which is why it’s important to serve as many institutions in the United States as possible. This rich diversity of institutions is one of the reasons why we built Plaid in the first place (and why getting U.S. coverage right remains our current priority). Without a way to connect with all these institutions from a single integration, it could take app developers years before they ever turn their attention to their own solution.

So when we talk about the number of institutions in the United States, there are a lot — and many more than we might find internationally. While there are other sources of personal financial data, like credit cards, any number much higher than 11,652 institutions is most likely counting duplicates (or triplicates, or quintuplicates… you get the idea). What’s more, not every one of these 11,652 and its subdivisions offers online banking, which is a prerequisite for connecting to a digital financial technology application. Take Carter Bank & Trust, for example, which has $4.5 billion in assets, but has yet to offer online banking.

Of course, while some institutions simply fail as businesses, others are acquired and have their depository accounts absorbed by peer and competitor institutions. Plus, institutions are constantly evolving their infrastructure, so a bank may completely change from an integration perspective even if it’s continuing to operate as it always was. Such reshaping of the financial industry obviously has big implications for us at Plaid, and for other financial infrastructure providers.

When we build an integration with an institution, we need to account for all of this history. Like it or not, a lot of the legacy infrastructure remains. (These stitches have occasionally been exposed to the consumer, too, even while the marketing and other operations of combined entities looks streamlined: Remember logging in to Bank of America online and choosing the state where the account was opened?) But in order to deliver the best user experience, these technical details have to be abstracted away. Imagine the burden on the consumer if Link actually exposed — and forced them to choose among — the 41 different Arvest Banks we consolidated in order to serve every Arvest customer. That’s why our work isn’t done if we’ve simply built integrations; we have to consolidate them in order to provide the best possible consumer experience. And that’s why we try so hard not to double-count them.

To be sure, building an inclusive financial ecosystem is a critical priority. But it turns out that identifying just how many institutions constitutes inclusive is an important exercise, too.