The Banker Who Knew Your Father's Name — And What Replaced Him

His name was something like Mr. Harmon, or Mr. Briggs. He wore a tie every day, even in August. He had a desk near the back of the First National Bank of whatever town you lived in, and he had been sitting at that desk for twenty-two years. When your father walked in to ask about a small business loan, Mr. Harmon didn't pull up a screen. He leaned back in his chair and said, "How's your mother doing after the surgery?"

That question wasn't small talk. It was the beginning of a credit evaluation.

Lending as a Local Act

For the first two-thirds of the twentieth century, most American lending — mortgages, small business loans, personal lines of credit — was handled by community banks and savings and loan associations with deep roots in specific towns and neighborhoods. These weren't branches of national conglomerates. They were local institutions, often founded by local families, staffed by people who lived nearby and had been doing business in the same zip code for decades.

The loan officer at one of these banks operated with a set of tools that would be unrecognizable today. There were financial documents, of course — income statements, tax returns, asset lists. But layered on top of those numbers was something harder to quantify: character assessment. How long had you lived in the community? Did you pay your local tradespeople on time? Had you honored previous commitments? Did people in town vouch for you?

This was, in many ways, a deeply human system. It was also, in ways worth acknowledging, a deeply flawed one. That same personal relationship model was used to exclude Black Americans, immigrants, women, and anyone outside the social networks of the men who ran these institutions. Redlining — the systematic denial of financial services to minority neighborhoods — was one of the ugliest expressions of relationship-based lending. The system's reliance on "character" often meant reliance on race, religion, and social class.

So the move toward standardization wasn't arbitrary. It was partly a response to real and documented injustice.

The Rise of the Score

Fair Isaac Corporation — the company behind what became the FICO score — introduced its credit scoring model in 1956, but the system didn't achieve broad adoption until the late 1980s and early 1990s, when Fannie Mae and Freddie Mac began requiring FICO scores for the mortgages they purchased. That decision changed everything.

The pitch was compelling on paper: replace subjective human judgment — with all its biases and inconsistencies — with an objective mathematical model that treated every applicant the same way. No more Mr. Harmon deciding he didn't like your handshake. No more loans going to golf buddies and being denied to people from the wrong side of town. Just numbers, applied uniformly.

In practice, the system delivered some of what it promised. Credit became more widely available. Lenders could process vastly more applications. The mortgage market expanded dramatically through the 1990s and 2000s, bringing homeownership within reach for more Americans than ever before — at least until the collapse of 2008 revealed that algorithmic lending at scale carried its own catastrophic failure modes.

But standardization also introduced a new kind of exclusion, less visible than the old kind but no less real.

What the Algorithm Can't See

The FICO score is built from five data categories: payment history, amounts owed, length of credit history, new credit inquiries, and credit mix. What it cannot measure — by design — is the texture of a person's actual financial life.

Consider a 28-year-old who graduated from college, spent five years paying off student loans aggressively, never carried a credit card balance, and has been steadily employed for four years. By the logic of a relationship-based lender, this person is an excellent credit risk. They've demonstrated discipline, stability, and the ability to manage debt responsibly.

But their FICO score might be surprisingly mediocre — because they have a short credit history, limited credit mix, and the rapid payoff of their student loans may have actually reduced their score by closing a major account. The algorithm sees the inputs it was designed to see. It cannot see the story.

This gap hits hardest for people who are new to the country, recently widowed, recently divorced, or who have simply managed their finances in ways that don't generate the kind of credit-bureau-visible activity the scoring model requires. An estimated 45 million Americans are "credit invisible" — they have no meaningful credit score at all — not because they're financially irresponsible, but because they've been operating outside the specific channels the scoring system monitors.

The Community Bank's Disappearing Act

The shift from relationship banking to algorithmic banking didn't happen in a vacuum. It was accelerated by the consolidation of the American banking industry, which has been one of the dominant financial stories of the past forty years.

In 1984, there were roughly 14,500 commercial banks operating in the United States. By 2022, that number had fallen below 4,200, according to FDIC data. The community banks that survived are often shadows of what they once were — operating within national compliance frameworks, using centralized underwriting software, and staffed by employees who rotate between branches and rarely build the multi-decade local knowledge that made relationship lending possible.

The big four banks — JPMorgan Chase, Bank of America, Wells Fargo, and Citigroup — now control more than 40 percent of all U.S. banking assets. Their loan decisions are made in underwriting centers, not local offices. The loan officer who might once have known your employer personally now feeds your data into a system and waits for an output.

The Middle Ground We Haven't Found Yet

The honest version of this conversation holds two things at once: the old system was personal but exclusionary, and the new system is consistent but blind. Neither fully serves the people it's supposed to.

Alternative credit scoring models — some using rental payment history, utility payments, and bank account data — are beginning to fill some of the gaps. Credit unions, which are member-owned and community-focused by charter, have maintained more of the relationship-banking ethos than commercial banks. Community Development Financial Institutions (CDFIs) specifically target underserved borrowers with more flexible underwriting.

But for most Americans, the primary experience of seeking credit remains a transaction between a person and a number — a number generated by criteria they don't fully understand, maintained by bureaus they don't control, and evaluated by institutions that have never heard their name.

Mr. Harmon had his biases. But he also had context. And context, it turns out, is something a three-digit score was never designed to carry.