The number on the application
You find the apartment, you like it, and you fill in the form. Somewhere between the form and the keys, a stranger you will never meet runs a check, and a single three-digit number comes back and settles the matter. If the number is high enough you are offered the lease, perhaps with one month’s deposit; if it is low, you are asked for two or three months up front, or a guarantor, or you are simply told the unit has gone. You were not in the room, you were not asked to explain, and the number spoke for you before you could. To most people that arrangement feels strange enough to sharpen into a question that has sent readers to this chapter: why does a three-digit score get to decide where I live.
The number is a credit score, and the same figure that gates the apartment also sets the interest rate on a car loan, feeds the premium on an insurance policy, while the record it is built from is, in many places, weighed before a job offer. It looks like a grade for your whole financial character, handed down from nowhere. It is neither of those things. It is the output of an industry, built up over more than a century, whose single purpose is to answer one narrow question about a stranger: if we extend credit to this person, how likely are they to pay it back. Following how that judgment was made, and then made into a number, is the way to see both what the score gives you and what it takes.
Before the number, a person decided
For most of the history of lending, the question of whether you could be trusted with money was answered by a person who claimed to know you. A local banker decided whether to grant a loan on the strength of your reputation, your standing in town, the look of you across the desk, and whatever he had heard. The decision was personal in the most literal sense: it lived in one man’s head, and it rested on his reading of your character. This had one great virtue and several matching faults. The virtue was judgment: a banker who knew your circumstances could see that a missed payment came from a bad harvest rather than bad faith. The faults were the reverse of that coin. The judgment was slow, it did not travel beyond the people who knew you, and it carried every prejudice of the man making it. Whole categories of person, women, newcomers, anyone of the wrong background, could be turned away at the door for reasons that had nothing to do with whether they paid their debts.
An organised trade in reputation grew up first around businesses, not consumers. When a merchant in one city wanted to sell on credit to a shopkeeper in another he had never met, he needed a way to judge a stranger at a distance, and agencies appeared to sell exactly that. The Mercantile Agency, founded in New York by Lewis Tappan in 1841, built a network of local correspondents who filed reports on the character and means of merchants across the country, and it grew into the firm later known as R. G. Dun and Company; a rival founded by John Bradstreet followed, and the two eventually merged into Dun and Bradstreet. These reports were narratives, not numbers. They described a man as honest and hard-working or as intemperate and slippery, in the moral vocabulary of the day, and their reporters, who in the early years included a young Abraham Lincoln, were recording reputation rather than measuring it. What mattered is that trust in a stranger had become a thing that could be written down, filed, and sold, which is the first step on the road to the number.
Trust becomes an industry
The century that followed turned that filed reputation into a manufactured score, and it did so in stages, each adding a piece the others did not have. The judgment that had once sat in a banker’s head was pulled out of it, spread across a records industry, handed to a statistical formula, placed under law, and finally compressed into one portable figure.
The first move was to do for ordinary people what the mercantile agencies had done for firms. Consumer credit bureaus appeared to keep files on shoppers and borrowers: the Retail Credit Company, founded in Atlanta in 1899 and now known as Equifax, sent investigators door to door and recorded in ledgers not only whether a person paid on time but details of their habits and reputation. By the middle of the twentieth century a thicket of local and regional bureaus held files on tens of millions of people, but the files still had to be read and weighed by a human clerk. The second move replaced that clerk with mathematics. In 1956 an engineer, Bill Fair, and a mathematician, Earl Isaac, founded a company in California, Fair, Isaac and Company, to sell lenders a statistical tool: a scorecard that took the facts on an applicant and returned a number ranking how likely they were to repay. Instead of a clerk’s intuition, a formula built from the records of thousands of past borrowers now estimated the risk of the next one.
The third move was law. For decades the files the bureaus kept were secret from the people they described, who could be refused a loan, a job, or an apartment on the strength of a record they could not see or correct. The Fair Credit Reporting Act of 1970 changed that, giving consumers the right to see their own file, to dispute what was in it, and to have errors investigated, and limiting who could pull a report and why. Four years later the Equal Credit Opportunity Act began forbidding lenders to discriminate, at first by sex and marital status and, after a 1976 amendment, by race, colour, religion, and national origin as well. Under the rules that implemented it, a scoring formula was barred from using those characteristics at all. The fourth and final move made the number universal. The bureaus and Fair, Isaac had produced many different scores for many different lenders, but in 1989 Fair, Isaac introduced its first general-purpose FICO score, sold through the credit bureaus and built to a single scale that ran from 300 to 850. For the first time a lender anywhere could ask for one standard number and get back the same kind of answer about anyone. The handshake had become a figure that any stranger could read.
What the number sees, and what it decides
To understand the score’s power it helps to be exact about the small thing it measures and the large number of things it is used to decide. A FICO score is built almost entirely from one source: the record of how you have handled borrowed money, as reported to the credit bureaus by your lenders. The company that builds it publishes roughly how much each part of that record counts. Your history of paying on time is the largest piece, about 35 per cent of the score; how much you owe relative to your available credit is next, about 30 per cent; the length of your credit history is around 15 per cent; and how much new credit you have sought and what mix of loans you hold make up the last 10 per cent each. What is striking is what the list leaves out. The score does not know your income, your savings, or your net worth. It does not know whether you are frugal or reckless with the money you are not borrowing. It rewards holding and using credit responsibly, which means a person who has always paid cash and never borrowed can have no score at all, and be treated by the system as an unknown rather than as safe.
That narrow measure is then put to a very wide range of uses, most of which have nothing to do with the loan it was designed to price. The same score, or a close cousin of it, is consulted when you apply for an apartment, when an insurer sets your motor or home premium through what the industry calls a credit-based insurance score, and when a utility decides whether to ask you for a deposit; and in many states, with your written permission, an employer weighing you for a job may be shown a version of your credit report, though not the score itself. A record of whether you repay debts has become a general signal of reliability, read across housing, insurance, and work by institutions that never lent you anything. The reach is the point, and it is worth seeing plainly.
What the number opened
Set against the personal judgment it replaced, the score’s first effect was to widen the door. When the decision to lend lived in a loan officer’s head, it travelled with all his blind spots, and a great many creditworthy people were refused for reasons that had nothing to do with their willingness to pay. The applicant who was the wrong sex, the wrong colour, or simply new in town and unknown to the bank could be turned away on sight. A statistical score, whatever its own faults, does not look up from the desk and see a face. Built from repayment records and forbidden by law from using race, sex, religion, or national origin, it judges the newcomer and the insider by the same rule, and it does so in seconds rather than weeks. A congressionally mandated review by the Federal Reserve in 2007 found that credit had grown more available over the previous quarter-century, and judged that credit scoring, by making underwriting faster and more consistent, had likely contributed to that wider access and to lower costs, while cautioning that the direct evidence on scoring’s own effect was limited. The person a local banker would never have met, and might never have trusted, could now be lent to on the strength of a record rather than a reputation.
This is the democratising half of the story, and it is real. Credit that once flowed through personal relationships, and therefore mostly to people who already had them, could reach anyone with a repayment history, wherever they lived and whoever they knew. The mortgage, the car, the card, and the small business loan became things a stranger could obtain from a stranger. A system that judges you by a formula can be a fairer master than one that judges you by a handshake, because the formula, at least, applies the same test to everyone it can see.
What the number keeps
The reach of that phrase, everyone it can see, is where the other half of the story begins. The same machinery that judges you consistently also remembers you permanently and follows you everywhere, and it can be wrong about you in ways no loan officer ever could. A banker’s bad opinion stayed in one town and faded when he retired. A credit file is national, durable, and portable: a missed payment, a disputed medical bill, a debt that was never truly yours can sit in the record for years and be read the same way by a lender in a city you have never visited. When the judgment lived in a person, its errors were local and mortal. When it lives in a file, its errors are lasting and travel with you.
Nor are those errors rare. The law grants the right to see and correct the file precisely because the file is often wrong. When the Federal Trade Commission studied the question, having 1,001 consumers review 2,968 of their own credit reports with expert help, it found that one in five consumers had an error on at least one of their three reports that the bureau corrected once it was disputed, and that for 5 per cent of consumers the error was serious enough to have pushed them into a worse pricing tier, so that they would have paid more for a loan or an insurance policy on the strength of a mistake. The right to dispute an error is not the same as the error being caught: the burden of noticing it, proving it, and chasing its correction across three separate bureaus falls on the person the mistake is quietly costing. Complaints about credit reporting are consistently among the largest single category the Consumer Financial Protection Bureau receives, which is what a durable, portable record looks like from the side of the person carried in it.
The exclusion has a second edge as well. Because the score is built from a history of borrowing, the person with no such history is not scored as safe but as invisible. The Consumer Financial Protection Bureau, which in 2025 corrected a widely cited earlier estimate, reckons that some 7 million American adults were “credit invisible” as recently as 2020, with no file at any of the major bureaus, and that tens of millions more hold files too thin to score, with those groups falling disproportionately on the young, the poor, and Black and Hispanic consumers. For them the system reproduces the old exclusion in a new form: not a banker’s prejudice, but a formula that has nothing to read, and so cannot say yes. That the cost of being outside the system, or of carrying its errors, tends to fall hardest on those with the least, and why being poor is itself expensive, is taken up where it belongs, in the volume on economic pathologies (Volume V, on why being broke is expensive). Here the point is narrower: the record that opened credit to millions is the same record that can shut a door on the basis of a mistake, or of a silence, and pass the same verdict on you in every city you enter.
A credit score measures how responsible you are with money.
Oversimplified Moderate confidence
There is a real grain of truth in the claim, which is why it is so widely believed. Paying your debts on time is a kind of financial responsibility, and it is the single largest thing the score rewards, so a high score is genuinely correlated with one important good habit. But the score measures that one strand and mistakes it for the whole cloth. It is built almost entirely from your record of borrowing, and it is blind to your income, your savings, and your net worth; it cannot tell a careful saver from a spendthrift so long as both pay their cards, and it can rate a debt-free person who has never borrowed as an unknown rather than as safe. It rewards using credit, not avoiding the need for it. So the score tracks how reliably you repay what you borrow, which is a narrow and specific thing, and not how responsible or how solvent you are, which is what most people hear in the word. The claim keeps the part that is true and quietly enlarges it into a measure of character the number was never built to be.
Sources
- Fair Isaac Corporation, “What’s in my FICO Scores” (myFICO) — the score’s stated composition (payment history ≈35%, amounts owed ≈30%, length of history ≈15%, new credit and credit mix ≈10% each), all of it a record of borrowing, with no input for income, savings, or wealth.
- Consumer Financial Protection Bureau, “Data Point: Credit Invisibles” (2015), with the Bureau’s 2025 technical correction — the original count of the credit-invisible (25.9 million in 2010) was roughly halved on correction to 13.5 million, about 7 million by 2020, with tens of millions more holding unscorable files, so a person who does not borrow is treated as unknown rather than as low-risk; the direct test of the “responsible with money” premise for the debt-free.
- J. Lauer, Creditworthy: A History of Consumer Surveillance and Financial Identity in America (2017) — the historical shift from character-based judgment to quantified scoring, and the argument that the score constructs a financial identity rather than reading a moral one.
- Confidence is moderate under the rubric: the pip scores the weaker of evidence directness and construct match, and construct match binds here. The evidence for what the score is built from is direct and published, but the claim turns on the fuzzy construct “responsible with money,” of which on-time repayment is one genuine component; the verdict is Oversimplified because the claim is partly true and overreaches, not because it is inverted or false.
Where the argument goes next
The three-digit number that decides your apartment, then, is neither a verdict on your worth nor a trick played on you. It is the latest form of a very old problem, how to trust a stranger with money, solved by pulling the judgment out of one person’s head and rebuilding it as an industry: a record of how you have borrowed, kept nationally, scored by formula, and read anywhere. That rebuilding opened credit to millions whom personal judgment shut out, and in the same motion it created a permanent, portable record that remembers a mistake as faithfully as a payment and carries both to every door you approach. The score gates your apartment because it is the cheapest way anyone has found to answer a question a landlord genuinely has, and its errors and its silences are the price the system charges for answering that question about everyone at once.
The credit score judges you one person at a time. The institutions that grew up around finance operate at another scale entirely, and they raise a different question about what all this machinery is finally for. The financial sector that holds the mortgages, prices the risks, and moves the savings has grown far faster than the economy around it, and the fees it takes are quietly clipped from the pension working on your behalf. What that larger and more expensive sector actually buys the people paying for it is the next chapter’s question.