You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.

I Worked at Capital One for Five Years. This Is How We Justified Piling Debt on Poor Customers.

The subprime lending giant is a textbook case in creating a corporate culture of denial.

The first thing you should know about a woman I know, who I’ll call Annie, is that she volunteers to sit at the hospital with people who are going to die alone, who have no family or friends to be with them during their last moments. “It’s obviously sad,” she told me, “but I feel like I have enough positive energy that I can share some.” And share she does: She cooks her coworkers’ favorite desserts for their birthday; she organizes anti-racism workshops and attends racial justice protests; she teaches ESL classes to recent immigrants. Annie is, in short, a very nice person. She works hard at being good, to be friendly and kind to everyone she meets.   

She also, for a time, made a living selling credit cards with high interest rates to people who were barely making ends meet.

Annie and I worked together at Capital One for three years. For a few months, I was her boss. I oversaw the bank’s “secured card” product—a credit card marketed to people whose credit is so bad they can’t get a credit limit of $300 at a 27 percent interest rate without putting down a security deposit. Ironically, at Capital One, the more of a positive-energy type you were, the more likely it was that you’d work in the subprime division. There, people like Annie and myself reasoned, the choices you made could, hypothetically, make things easier for struggling families. We told ourselves that such families likely didn’t have any better lending options. And for poor, under-banked households, many lending options are far worse than Capital One.

The real question, of course, isn’t whether a credit card with a 27 percent interest rate and a $39 late fee is better than a payday loan. It’s whether Capital One’s marketing campaigns push people into debt who would have otherwise avoided it; whether it is actually in a person’s best interest, desperate though they may be, to borrow money at an exorbitant rate; and whether this enterprise is ethically defensible—in particular, for the decent, hard-working employees who toil every day to make Capital One’s mercenary strategy a reality. Because the ugly truth is that subprime credit is all about profiting from other people’s misery.

In 2012, the year I started my first Capital One internship, the company’s acquisition of HSBC’s credit card business went through, making it one of the largest subprime credit card issuers in the U.S. The decision to double down on those Americans struggling to get by has paid off handsomely. 

The credit card titan’s newly-constructed 31-story glass headquarters in McLean, Virginia, is but one lavish testimonial to the success of its bottom-feeding business model. Capital One collects $23 billion in interest per year—an average that works out to $181 from each family in America. Of course, not every family has a Capital One account, and most public surveys say roughly half of people with credit cards pay them in full and accrue no interest. So simple math tells you that many families are paying Capital One at least $800 in interest every year. 

And most of that interest gets paid by the families who can least afford it. According to data from a 2018 Federal Reserve survey, people who report an unpaid credit card balance “most or all of the time” were nearly five times as likely to describe themselves as “struggling to get by” or “just getting by” than the people who paid their credit card bills in full every month. They were nearly 50 percent more likely to have an income under $50,000, 2.5 times as likely to describe the economic conditions in their community as “poor,” and three times as likely to have skipped prescription medicine or doctor’s visits because of cost.

People at Capital One are extremely friendly. But one striking fact of life there was how rarely anyone acknowledged the suffering of its customers. It’s no rhetorical exaggeration to say that the 3,000 white-collar workers at its headquarters are making good money off the backs of the poor. The conspiracy of silence that engulfed this bottom-line truth spoke volumes about how all of us at Capital One viewed our place in the world, and what we saw when we looked down from our glass tower. This is not meant to offer a broad-brush indictment of business at Capital One; it is hardly the only corporation that has been ethically compromised by capitalism. It is, however, meant to shine a few photons of light on the financial industry in a post-crisis age of acute inequality.  

Sometimes at Capital One, you would be working on a Powerpoint slide that showed “chargeoff rates” increasing. Your slide would feature a bunch of colored lines: one showing that ten in 100 people who opened this type of credit card failed to repay their debt within the first year; the next illustrating how the same thing happened to 15 in 100 people the following year. A curious colleague might walk by and say things like, “Oh, wow, can I take a look at those curves?” Said colleague might then offer up some comment pertaining to the work they did downstream from these indebtedness trends: “That’s fascinating. Is this deseasonalized? Does the dollars-bad chart look the same as the people-bad chart?”

For the 10-15 percent segment of people represented on the graphs, the thought experiment was far less abstract—and substantially less pleasant. Their credit scores would have tanked by 100 points; they would be fending off increasingly urgent calls from debt collectors. Capital One did a great deal of its own debt collection instead of outsourcing it, and we patted ourselves on the back for that, since it was considered more humane. Chances were good that Capital One would eventually sue these non-payers, with the hopes of garnishing their paycheck to get that money back, with interest.

Amid the daily office banter at Capital One, we hardly ever broached the essence of what we were doing. Instead, we discussed the “physics” of our work. Analysts would commonly say that “whiteboarding”—a gratifying exercise in gaming out equations on the whiteboard to figure out a better way to build a risk model or design an experiment—was the favorite part of their job. Hour-long conversations would oscillate between abstruse metaphors representing indebtedness and poverty, and an equally opaque jargon composed of math and finance-speak.

If you were not familiar with the almanac of metaphors—many of which, as I understand it, were specific to Capital One—you would not follow the conversations. The “bathtub,” for example, denotes a loan portfolio, because it’s like water down the drain when you lose customers—either because they have closed their account or were fed up with Capital One or have involuntarily defaulted on their loan. When you spend tens of millions of dollars on marketing, that’s turning on the spigot for new water in your “bathtub.”

It was common to hear analysts say things like, “I just love to solve problems.” But what they were really doing was solving something closer to puzzles. It’s clear to me, for example, that the janitor at my middle school solved problems when she cleaned up trash. It’s far less clear whether analysts at Capital One are solving problems or creating them. In either event, the work culture at this well-appointed lender of dwindling resort is pretty much designed to encourage former students of engineering or math to let their minds drift for a few years and forget whether the equations in front of them represent the laws of thermodynamics or single moms who want to pay for their kids’ Christmas gifts without having to default on their rent or utilities payments.

Capital One is a very young company, particularly relative to other banks. It was founded in 1988 by the all-too aptly named Rich Fairbank, who came from the world of consulting. Virtually everyone I worked with was 45 or younger. Most were under 30. By the age of 25, I was what you might call a “middle manager,” with a team of analysts working for me and day-to-day responsibility over a substantial part of the business. 

My trajectory wasn’t all that unusual. The company was willing to put, for better or worse, extraordinary amounts of trust in people very early in their careers. For the many young people at Capital One, the rampant euphemisms they used were not a defense mechanism. Bad feelings rarely impinged on the sunny state of mind that typically grew out of their successful lives, so they did not need to mount a defense.

Before I managed Capital One’s secured card product, I worked on what we called “Mainstreet proactive credit limit increases” or “Mainstreet pCLIP” for short.  Mainstreet was yet another piece of euphemistic in-house jargon; it meant subprime. As for proactive credit limit increase, it meant raising the cap on how much someone is allowed to borrow—without getting their permission to raise the cap. 

The emails we used to send these “Mainstreet pCLIP” customers would go as follows: “Elena Botella, you’re a valued customer, and we want you to get more out of your card. So recently, your credit line was increased to $6550.00. This gives you more in your wallet, which gives you more flexibility. Thank you for choosing Capital One®. Enjoy your higher credit line.”

At any bank, if you have a low credit score, you’re only likely to get a credit limit increase if you’re getting close to your existing credit limit. So if you got that email, you probably had a few thousand dollars of Capital One credit card debt at an interest rate of at least 20 percent. That implies you were probably paying Capital One around $40 in interest per month or more. You might want or need to borrow more money on top of what you’ve already borrowed, but I always thought it was a little bit sick for us to be telling people to “enjoy” their higher credit line. It felt more than a little like shouting, “Enjoy getting into more debt, suckers!” before disappearing in a cloud of smoke and speeding off in a Tesla.

In my coworkers’ defense, nearly everyone agreed with me that repeating “more...more...more” in that email was redundant, if not manipulative, and we eventually changed the text. But the pitch still represented an irreconcilable gap: While the recipients of the email might, at best, regard it as one additional debt instrument to deploy for the next financial emergency lying in wait, Capital One employees must have imagined that borrowers were positively excited to take on more debt as soon as possible. 

Sadly, this latter view was not completely unreasonable. As Scott Schuh and Scott Fulford have shown in a paper for the Federal Reserve of Boston, people who get credit limit increases tend to keep their “utilization” constant. In other words: If a person is carrying a $1,500 balance when they have a $3,000 credit limit, you’d expect them to start carrying a $4,000 balance if the limit is raised to $8,000. If most people use the full credit-limit increases they are offered, the thinking goes, that must mean that most people want to borrow more money. If you lend them more money, you are “meeting customers’ needs.” 

Because the borrower’s pain was not at the forefront for analysts, lingo like “pBad” (the percentage of people who can’t repay their loans), “second-order risk” (when customers who would have been able to repay a small loan default because they borrowed more than they could handle), “flow rates” (the percentage of people who will miss the next payment), “HBRs” (high-balance revolvers, or people who have a lot of debt) is not analogous to a military planner referring to “collateral damage” to talk about dead civilians. It is far more abstract. 

Capital One’s culture of experimentation also acted as a kind of buffer. Fast Company has reported that Capital One runs 80,000 experiments per year. As Christopher Worley and Edward Lawler III explain in the journal Organizational Dynamics, a bank like Capital One can randomly assign differing interest rates, payment options, or rewards to various customers and see which combinations are most profitable for any given segment of people. It’s not so different from how a pharmaceutical company might use a randomized control trial to test whether a new drug is effective, except that the results of the bank’s experiment will never get published, and instead of curing diseases, the bank is trying to extract more money from each customer. The use of experiments is itself an act of psychological distancing; it allows the analysts controlling the experiment to resolutely apply its findings as a profit-maximizing mandate without giving the strategy a name such as, oh, “predatory lending.”

In this faintly lab-like register of business rhetoric, it’s not necessary to say things out loud like, “We should have a credit card where people put down a $99 security deposit to get a $200 credit limit, paying hundreds or thousands of dollars of interest to Capital One, even though many of those people will default.” Nor would you have to say, “This person has maxed out every credit limit increase we’ve given them in the past on some stuff they probably needed like new tires, but also on other stuff like concert tickets. But since they always make the minimum payment it’s probably safe to raise their credit line so they’ll take on more debt, because as long as they keep being able to afford making the minimum payment for another two years it’s actually okay if they eventually can’t repay us.” 

The rise of data science, machine learning, and artificial intelligence means that you don’t need venal corporate tycoons wearing Monopoly Man hats to grind the faces of the poor into the dirt. Under the data-driven directives of Capitalism 2.0, you can have a bunch of friendly data scientists who don’t think too deeply about the models they’re building, while tutoring low-income kids on the side. As far as they’re concerned, they’re refining a bunch of computer algorithms.

Executives at Capital One tend to get their questions answered. There were many nights I ordered pizza to our office for dinner so that I could hunker down for another five hours to make slides for a “Rich Deck.” This was the Capital One term of art for the very special kind of PowerPoint that CEO Rich Fairbank would see after a few people more senior than me had the chance to massage it into shape. If Fairbank cared to know the answer to any question—such as, “How many of the loans that we give out actually make the borrower’s life better?” or “What are the consequences of raising our credit card interest from the prime rate plus 19 percent to the prime rate plus 23 percent on child hunger in America?”—he could have gotten thoroughly researched answers. But those are the kinds of questions that the entire Capital One workplace was designed to drive out of view.

Still, Capital One was at least aware that it needed, at least every once in a while, to shore up its reputation and make its employees feel good about their occupation. When executives wanted to inspire their underlings to savor the broader social value of the work they were doing, under the banner of the “mission” to “change banking for good,” the most common strategy was to introduce a Powerpoint slide comparing a credit card to alternatives. A payday loan has an APR of 400 percent. One of Capital One’s competitors, First Premier, charges a 36 percent interest rate on top of a one-time set-up fee of $75, on top of more than $100 in additional monthly and annual fees, all to gain access to a $400 credit line.

It’s hard to say whether the world would be better off without credit cards. It’s true, of course, that sometimes people really do need to borrow money. I talked to one woman in Virginia who borrowed money on her credit card right after she graduated college to help pay her rent and buy her groceries until she found a job. She said her parents lived in the middle of nowhere, with grim prospects for employment in the knowledge economy. Borrowing money appreciably improved her life prospects for the better.

But the morality of the credit card executive is a morality of autonomy: If I’m giving people choices, that’s a good thing. “How could I be making this customer worse off,” a credit card executive asks herself, “if nobody is forced to use this product?” The way she goes to sleep at night is by imagining a quasi-mythical customer who would have lost her job if she couldn’t have put the new tires for her car on her credit card. Whether that person resembles one in every two customers, or one in every 50, or one in one million, is conveniently elided, despite the immense research resources at her disposal.  

When I was at Capital One, I wanted to understand if it was possible to keep loans as an option for the people who have exhausted all their better alternatives—without also causing suffering for those who would be better off forgoing purchases or borrowing money from friends and family. After five years, I concluded it was more or less possible to achieve that goal—to do the good loans without doing the bad loans. I also realized something that will sound obvious to a reasonably smart adult who doesn’t work at a credit company: an ethical corporation could be tempted by compelling evidence about the suffering it caused to relinquish some of its massive revenue. But over the long run, a publicly traded company wasn’t going to sacrifice a meaningful amount of income to avoid destroying lives—unless the law required it.

At the very least, I wanted my coworkers to swim in the pain with me, to think about it when we ate açaí bowls in the cafeteria, to foreground it before we made every decision. But that’s not how glass towers work.