The History of the Credit Card

A single card payment involves at least four parties: the person using the card, the store selling the goods, the company contracted with the store to process the payment, and the company that issued the card. The last two are usually different companies, meaning that what we casually lump together as "the credit card company" actually encompasses three distinct types of businesses.

To take an example straight from a report by the U.S. Government Accountability Office, it looks like this: when a customer pays $100, the merchant pays a fee of $2.20. Of that, $1.70 goes to the company that issued the card, and $0.50 remains with the company that processed the payment. What the merchant actually receives is $97.80. The customer paid $100, and the merchant received $97.80. The difference was split where the customer could never see it.

There is one more branch to this. If the customer does not pay off the full balance this month and rolls part of it over to the next, interest is charged on that amount. From that single transaction, you now have two streams of revenue: the fee paid by the merchant and the interest to be paid by the customer. This series is ultimately the story of when these two streams were created, and which of them grew larger.

The structure itself is familiar. Credit card companies do not manufacture goods. They connect sellers with buyers and take a fee in the middle. It looks just like what we call the platform economy today. The only difference is that in the 1950s, this platform ran on paper and postal mail—with neither payment terminals nor telecommunication networks, relying solely on printed booklets of participating merchants and bills mailed at the end of each month.

That is why there is one question we must keep asking throughout this story: at this very moment, who is paying how much to whom? Even as cards shifted from paper to plastic, and then onto smartphone screens, the answer to this question changed only a few times. And each time it did, the entire playing field shifted dramatically.

Yet the most famous story about how this whole thing began mentions none of what you just saw. Let's start by taking a close look at that scene.

The story goes like this. Around 1950, a businessman dining at a New York restaurant realized he had forgotten his wallet only when the check arrived. Deeply embarrassed, he resolved never to experience such a thing again and created a card. That company was Diners Club.

It is a neat story. Yet in our research, we found no contemporary records to support this scene. Instead, evidence pointing in the opposite direction turned up. Matty Simmons, the first employee and press agent for Diners Club, later wrote in his book that it never happened and that he had made up the story. He explained the reason, too: he needed to make the birth of the card scheme sound compelling.

So, it is most accurate to frame this part like this: No one has proven that the anecdote was physically impossible. What is confirmed is the fact that the very person who spread the story admitted he made it up. The weight of evidence drops drastically right here. It makes more sense to read this anecdote simply as a legend passed down over the years.

Why, then, did this story survive? Because the actual surviving records are boring. All they did was set up a company, compile a list of restaurants willing to accept the card, and mail that list to members. What the publicist needed was a single scene to mask all that boredom. And that scene is still featured on the official Diners Club website today. The company even commemorates that return visit in February 1950, calling it the "First Supper."

That scene does one more thing. It frames the card as an idea born out of one person's personal inconvenience. That is the exact reverse of the framework we saw in the previous chapter. In reality, what came first was not the inconvenience, but the business model. The math of taking a transaction fee from merchants was already set, and all that remained was recruiting merchants to accept it. Well-crafted origin stories usually flip this order upside down.

Where, then, did that business model come from? To understand that, we need to look at what existed before cards were ever introduced.

Buying on credit is far older than cards. Stories are widely told that around the 1920s, department stores and gas stations began issuing metal plates or paper cards to their regular customers. The department store versions were called charge plates, while gas stations called theirs courtesy cards.

Yet these all shared one common trait: they could only be used at the specific shop or company that issued them. You couldn't use a department store card to buy a meal at the restaurant next door. It was a closed loop.

What changed in 1950 was that this loop opened up. In a report to Congress, the Consumer Financial Protection Bureau noted that in this year, Diners Club was the first to create a card accepted across multiple merchants by partnering with local restaurants. Standing in the middle as a third party—neither the buyer nor the seller—it tied completely unrelated businesses together under a single card. The very structure we saw in the previous chapter was built here for the first time.

Yet Diners Club wasn't the only one to think of this structure. According to collection descriptions at the Smithsonian's National Museum of American History, the culinary magazine Gourmet launched a card using the exact same model around 1954. The booklet distributed with the card listed 600 hotels and restaurants that accepted it. The magazine Esquire followed with its own version starting in 1956. Magazine publishers had entered the card business. Since they already held lists of subscribers and advertisers, it was perhaps only natural.

The story ended in acquisitions. According to the same museum notes, American Express acquired Gourmet's club in 1958, while Diners Club bought Esquire's club. It wasn't invented by a single player; multiple contenders jumped in at once, and the bigger players survived.

This is where the heart of the story begins. Revenue for early cards came not from the cardholders, but from the merchants.

The Smithsonian's National Museum of American History describes Diners Club’s operation like this: they billed members at the end of the month and sent restaurants the amount minus a 5–7% processing fee. That was the price restaurants paid in the United States in the 1950s. Diners paid the exact face value of their meals; the ones taking the haircut were the restaurants. Of the two revenue streams we looked at in the previous chapter, only the merchant side existed back then.

Why did restaurants accept this? Because they believed card-carrying customers would boost their sales. And that was precisely what the card company was selling: a member directory. From the card company's perspective, the actual product in this business wasn't a piece of plastic, but the information that "here are people carrying this card."

Thanks to this structure, companies other than banks could enter the card business. Since it worked by paying on behalf of customers and collecting at the end of the month rather than lending money, there was no need to take deposits or underwrite loans. All they needed were two lists: a list of merchants willing to accept the card, and a list of people who found that directory valuable. Here lies the answer to why restaurant charge cards came before bank cards. It is also why a magazine publisher jumped into this business.

However, there was a tricky problem with this business from the very start. If few merchants accepted the card, people wouldn't carry it; if few people carried the card, merchants wouldn't accept it. Neither side wanted to make the first move. That's because it was something that only works if you gather both sides at the same time. What economists call the network effect looks exactly like this.

This is why early card companies mailed a booklet of participating merchants along with the card. The 600-venue list included with the Gourmet card was not just a courtesy; it was proof. Proof that there were plenty of places to use it. And as the scale grew, this challenge grew even heavier. This is precisely why banks, despite fiercely competing with one another later on, ended up building their card networks together.

In 1958, credit cards split down two paths.

That year, Bank of America rolled out BankAmericard in California. The Smithsonian's National Museum of American History records this as the first general-purpose credit card and the first card that let you pay over time instead of paying the full balance at the end of the month. On the outside, it looked much like earlier cards, but the engine inside was completely different.

Cards like Diners Club required you to pay off whatever you spent that month by the next bill. You postponed the payment date, but you did not carry debt around. The new card allowed you to pay just a fraction and roll the rest over to the next month. Interest was charged on the rolled-over amount. That is the feature we know today as revolving credit. If you could not clear it the next month, interest was charged again on top of the balance that already had interest added to it.

This single difference divided the scale of the business. A card paid in full brings in only merchant fees from a transaction. A card paid over time brings in both merchant fees and interest from that same transaction. Diners Club fell behind not because it was a bad card, but because it was a card that left no debt behind. That was the architecture the company deliberately chose from the beginning, and back then, it was considered the safe design.

When banks entered the fray, another hurdle emerged. There was a hard ceiling on how many merchants a single bank could sign up alone. So starting in 1966, Bank of America began licensing its card to other banks, and that same year, several New York banks formed a cooperative called Interbank. The former became Visa in 1976, and the latter became Mastercard. Banks that competed fiercely with each other decided to build out the card network together.

The race for market share got reckless. Records from the Federal Reserve system note that some banks, determined to pull ahead, launched sweeping mass campaigns and mailed cards to consumers without their consent. Cards arrived in mailboxes without anyone applying or undergoing a credit review. When the U.S. Truth in Lending Act was first amended in 1970, this practice was outlawed, and a cardholder's liability for unauthorized charges on a lost card was capped at $50. The reckless scattershot approach ended up summoning the rulebook.

The moment interest becomes revenue, laws governing interest begin to determine the scale of the business.

In the United States, each state had its own interest rate cap. Credit cards, however, traveled across state lines through the mail. If a bank based in Nebraska sent a card to someone living in Minnesota, which state's cap should apply? This question sparked an actual lawsuit.

On December 18, 1978, the U.S. Supreme Court delivered the answer in Marquette National Bank v. First of Omaha Service Corp. Under a federal statute allowing a national bank to charge interest rates permitted by the laws of the state where it is located, the Court held that the Nebraska bank could apply Nebraska's permitted rates even to cardholders in Minnesota. That bank was charging an annual rate of 18% up to $999.99 and 12% above that amount, whereas Minnesota's cap was 12%.

The logic was straightforward: where a bank is 'located' is determined by its chartered address, not by where the customer lives or where the transaction took place. The statutory clause itself came from an old law, and the Supreme Court simply applied that existing rule directly to credit cards.

The impact was immense. It is widely viewed as a turning point that effectively altered the leverage that state-level caps held over credit card rates. That said, it is difficult to assert that this single ruling alone dictated subsequent interest rate levels. The annual percentage rate printed on a card—namely, the nominal interest rate—is not a number set by card issuers alone. The U.S. Consumer Financial Protection Bureau reports that between 2022 and 2024, most of the increase in U.S. credit card interest rates stemmed from changes in the prime rate, which serves as the benchmark for variable interest rates. A credit card rate is a figure shaped together by card companies, benchmark interest rates, and the law.

So, where do those numbers stand today?

The US Consumer Financial Protection Bureau submits a report on the credit card market to Congress every two years. A 2009 law mandated it that way. Thanks to that, this is one of the few markets where the numbers are out in the open.

Looking at the 2023 report, the largest revenue category for general-purpose cards at large US banks was interest income. As of 2022, it stood at 12.3% relative to average card outstandings. Next was interchange—the share of fees paid by merchants that makes its way to the card issuer—at 7.9%, followed by various fee income at 1.5%. All three are measured against the same denominator.

Step one level deeper, and the underlying structure comes into view. The same report notes that of that interchange income, nearly nine in ten dollars flowed back out as points, miles, and cash back. It also tallies that cardholders who carry a balance to the next month paid almost all of the interest and three-quarters of the fees, while receiving less than 30% of the reward dollars. These are figures from the US, for 2022, as tallied by this agency.

So it boils down to this: The money paid by merchants mostly funds rewards, and the bulk of the remaining profit comes from interest. Merchant fees, which were the card's very first source of revenue, did not disappear, but their role has shifted.

That fee itself is no simple figure, either. According to an analysis by the US Government Accountability Office, the number of domestic credit card interchange rate tiers in 1991 was 4 each for Visa and Mastercard, but by 2009 had expanded to 60 for Visa and 243 for Mastercard. This was the result of rates fragmenting across merchant types, card tiers, and transaction methods. A rate is not a single number; it is an entire matrix.

Let's point out one more thing. It is easy to assume that the card networks pocket that fee, but Visa describes itself like this in its annual report: it is not a financial institution, does not issue cards, and does not set the interest rates or fees charged to cardholders. It does set the default interchange rates applied when there is no separate agreement, but it also explains that the fees it collects do not come from that pool. The place that sets the rules and the place that gets the money are different.

To add just the sheer scale in numbers: in 2024, US card purchase volume was $3.6 trillion, balances exceeded $1.2 trillion, and interest assessed to consumers was $160 billion. The average APR on general-purpose cards was 25.2%, and about half of accounts carried a balance. The same report also looks at this market broken down by credit score tiers, noting that nearly all of the spending growth in recent years came from the higher-score side.

What popularized credit cards in South Korea was not competition among banks, but the tax system.

The beginning looked familiar. According to records from Woori History Net, the forerunner of Korean credit cards was a store card issued by Shinsegae Department Store in 1969. It was a card you could use only at that department store. Kookmin Card arrived in 1980, followed in 1982 by BC Card, which multiple banks shared, and debit cards made their debut in February 1996.

The paths diverged in the late 1990s. According to records from the National Archives of Korea, the monthly limit on cash advances was abolished in May 1999, and an income tax deduction for credit card spending was introduced in August of that year. In January 2000, tax credit benefits were expanded for business owners who issued card sales slips. Consumers were given an incentive to save on taxes by using cards, while merchants were simultaneously given an incentive to save on taxes by recording card sales.

The response was swift. According to the same records, between 1999 and mid-2002, the number of credit cards jumped from 39 million to 100 million, and credit card transaction volume relative to nominal GDP surged from 9.3% to 94.1%. In just over three years, it had multiplied more than tenfold.

Then came the reckoning. The one-month delinquency rate rose from 2.6% at the end of 2001 to 6.6% at the end of 2002, and reached 14.1% at the end of 2003. The number of people with delinquent credit jumped from 2.45 million in 2001 to 3.72 million in 2003. Failing to pay back as promised—in other words, default—came crashing in all at once.

Here, the story did not end as a crisis confined to credit card companies. Because card issuers do not accept deposits, they raise money by issuing bonds. Other financial institutions held those bonds. That is why the government's measures announced on April 3, 2003, included extending the maturity of 44.5 trillion won in credit card bonds, establishing a 5.6 trillion won dedicated investment fund, and recommending a 4.6 trillion won additional capital injection by controlling shareholders. This was not about saving individual firms, but a measure to prevent the bond market from freezing up. It was a defining moment showing what systemic risk truly means. By the end of 2004, the delinquency rate had dropped back to 9.0%, and the return on total assets recovered from -20.3% to 3.9%.

Perhaps because of that experience, South Korea today does not leave merchant processing fees entirely to the market. Following an announcement by the Financial Services Commission on December 17, 2024, the preferential credit card fee rate for micro-merchants with annual sales of 300 million won or less was reduced from 0.50% to 0.40%, and for the 300 million to 500 million won bracket, from 1.10% to 1.00%. Covering approximately 3,046,000 merchants, this took effect on February 14, 2025. In principle, the standard cycle for recalculating merchant fees was also adjusted to every six years.

There is no need to try to predict the future. Instead, guided by the rules this story has revealed, let's look at what signals can help you spot shifts in the game before anyone else.

First, where the money comes from. From the very beginning, revenue for this card has been split into two streams: merchants and cardholders. When regulations hit one stream in a country, the other stream moves in response. Countries like South Korea, which recalculate merchant fees on a regular schedule, strike a different balance than countries like the US, where fee schedules are splintered into hundreds of tiers. When reading news about cards, checking which of these two streams is involved will keep you from getting confused.

Second, where rewards come from. US data shows that a significant portion of the fees paid by merchants goes right back out as points and cash back. If regulations push those merchant fees lower, rewards are bound to shift alongside them. At least among major US bank cards, these two figures have historically moved hand in hand.

Third, the share of people who carry a revolving balance. In the US, this share was about half in 2024. Because this number tracks both the economy and interest rates, it often moves earlier than headlines about delinquency rates.

Fourth, payment methods growing outside card networks. If direct account-to-account transfers or alternative payment tools expand, the very foundation of those two streams—merchant fees and interest—begins to wobble. Still, it is too early to jump to conclusions here. According to Federal Reserve surveys, in 2024, the breakdown of US consumer payments by number of transactions was 35% credit cards, 30% debit cards, and 14% cash. In fact, cash payment frequency has held steady at 7 times a month since 2020. One has never erased another yet.

Finally, just one more thing. Throughout this story, what shifted the card's place in the world usually came from outside the card industry itself: a magazine's subscriber list, licensing agreements between banks, laws sparked by mailing unsolicited cards, a court ruling over which state's interest cap applied, and tax deduction schemes. Whatever changes this object next will likely come from somewhere outside the news about cards. Where that will be is well worth watching.