Why Does Every Card Reader Ask for a Tip Now?

Why do card readers ask for a tip? Because the suggested amounts move what people pay, and that has been measured on real payments. Nobody has published why businesses switched the prompt on. In one New York taxi study of over 13 million rides, changing the suggestions raised tips by more than a tenth¹.

Do suggested tip amounts change how much people tip?

Yes, in the best data available. In 2009 New York taxis showed riders a screen with three buttons¹. Below a fare of $15, the buttons offered $2, $3 or $4¹. At $15 and above, they offered 20, 25 or 30 percent¹. Economists Kareem Haggag and Giovanni Paci compared riders just either side of that line, across more than 13 million rides, and found that those who saw the percentage buttons tipped about 27 to 30 cents more, over a tenth of the average tip at that point¹.

Nothing forced anyone to press a button. Riders could type any amount or none¹. So the screen shifted behaviour without taking a choice away.

Why it works is not settled. The authors say they cannot separate a few explanations: riders reading the buttons as what is normal, not wanting to fall below it, or simply taking the easy option¹. The evidence covers what the screens do. It does not say why any business turned one on.

Does typing your own amount get you out of it?

Apparently not. A 2026 working paper looked at two New York taxi card machines that showed identical buttons of 20, 25 and 30 percent, but applied them to different totals, so the same percentage meant more money on one of them². Riders who pressed a button paid more on that machine, as expected. The surprise was the riders who ignored the buttons and typed their own figure: they also paid more, by roughly 8 to 10 percent².

The reading is that the screen sets a sense of normal that reaches even people who refuse its options. This is one working paper, not yet replicated, so it deserves a light grip².

It also sits awkwardly beside what people say about themselves. In a Pew survey of 11,945 US adults, 77 percent called the quality of service a major factor in how much they tip³. That is what people report, and they mean it. But the two taxi machines were not providing different service. They were displaying different numbers².

Does the same screen work when nobody is watching?

Much less. In 2017, when Uber had only just added in-app tipping, researchers randomly showed more than 12 million riders different suggested amounts⁴. Here the rider tips on their phone after the ride, with the driver already gone. The highest suggestions lifted the average tip by only about 2.5 percent over the lowest, against more than 10 percent in the taxis⁴.

The authors’ reading is that suggested amounts and a watching person work together: the number tells you what normal is, and someone right there makes falling short of it feel like something⁴. That is an interpretation, not a controlled test. A taxi and an app differ in other ways too⁴.

With no one watching, most riders simply did not tip. About 16 percent of trips were tipped, and about 60 percent of riders never tipped once in four weeks⁵.

A separate study points the same way from the customer’s side⁶. Across a field study and four experiments, once digital checkout let the worker see the tip, customers said they were less likely to return or recommend the place, and felt less generous and less in control. Letting them change their first choice softened the effect⁶. Only the abstract was available, so there are no figures to quote.

Nobody has run these tests at a coffee counter with a barista and a queue. That is where the findings point, not where they were measured.

Do higher suggested tips backfire?

Partly. In the same taxi data, higher suggestions raised the average tip, but they also raised the share of riders who left nothing, by 1.7 and 2.8 percentage points in the two comparisons⁷. That is small in absolute terms, two or three more riders in every hundred, and large in relative terms: riders were over 50 percent more likely to tip zero⁷. The authors speculate about the reasons, and the study did not test them⁷.

The pushback is visible in surveys as well. Asked about businesses suggesting tip amounts on a bill or screen, 40 percent of US adults said they oppose it, 24 percent favour it and 32 percent said neither³. Some 23 percent called social pressure a major factor in whether they tip³. These are opinions, but they show the screen divides people, and the person behind you in the queue may feel the opposite of you.

Why does tipping feel so different at a counter?

People say it is different. In the Pew survey, 92 percent of people who eat at sit-down restaurants said they always or often tip, against about a quarter at coffee shops and 12 percent at fast-casual counters, where 45 percent said they never do⁸. Those are self-reports, and they describe a mismatch between the screen and the social script for a counter, not a personal inconsistency.

And 72 percent of US adults said tipping is expected in more places than five years ago⁹. That is a perception, not a count of screens⁹.

Is tipping culture actually new?

The screens are new. The argument is not. Tipping spread through the US after the Civil War and drew resentment from the start, and between 1909 and 1915 six states made it a crime¹⁰. Every one of those laws was repealed or struck down by the 1920s¹⁰.

What the research doesn’t show

It doesn’t show why businesses switched their prompts on. No source read here records that, and claims that screens were designed to induce guilt are not backed by any study of intent¹.

It doesn’t show that people feel guilty. The taxi and Uber studies measured payments, not feelings. “Guilt” here is the everyday word for a pause many people report, and the closest measured pieces are the authors’ reading that norms and a watcher work together⁴ and a self-reported social-pressure share of 23 percent³.

It doesn’t show the effect at a coffee shop counter, or anywhere outside the US. The typed-amount result is from a working paper and the Uber comparison is an interpretation²⁴. The visible-tipping study is an abstract without figures⁶.

Nothing here says whether, or how much, anyone should tip.

So what does the pause over the buttons mean?

The hesitation at the reader is not a measure of generosity or of character. On the evidence, it is what happens when a screen shows a number that stands in for normal¹ and, at least as the researchers read it, a person is standing there to see whether you match it⁴. The buttons, the number you type, the face across the counter and the grey button at the bottom all shape the outcome. None of them is a test of who you are.

If you are the person who feels every payment before it even happens, that is a different question, and a measured one: why some people can’t spend money even when they can afford it.

Educational content, not financial advice.