If you have recently been in a taxi in New York City, you may have noticed a credit-card touchscreen, which suggests three possible tips. For rides of more than $15, the suggested amounts are usually 20 percent, 25 percent or 30 percent. You can give a larger tip, a smaller tip or no tip at all, but it’s easiest just to touch one of the three conspicuous options.
The touchscreen makes everything simpler. It also raises an intriguing question: Do the suggestions affect the average tip? Behavioral economists would offer a clear prediction: Because people don’t like to do even a little bit of extra work, the suggestions will matter a lot, and the average tip will increase significantly.
The prediction has turned out to be right, with one important qualification: There’s a backlash effect, with more customers giving no tip at all. This natural experiment illuminates human behavior in a lot of diverse settings, and it has implications for business and for public policy as well.
The instructive study has been done by Kareem Haggag of the University of Chicago and Giovanni Paci of Columbia University, who have compiled data on more than 13 million New York taxi rides. Using the standard social-science jargon, Haggag and Paci describe the suggested percentages as “defaults,” in the sense that they establish what customers will do if they don’t exert extra effort.
To test the effect of the defaults, they examine two data sets. First, some companies provide somewhat lower defaults of 15 percent, 20 percent and 25 percent. Do people tip less when presented with these lower defaults? Second, companies with the higher defaults typically provide lower default percentages for smaller fares (less than $15). Do those lower percentages reduce tips?
The main finding is that the higher defaults lead to a significant increase in tips -- by an average of more than 10 percent. If a driver makes $6,000 in tips in a year, the higher defaults produce a nice $600 raise -- and a big increase in total revenue for New York City taxi drivers as a whole.
Why, exactly, do the defaults matter? Haggag and Paci explore three explanations. The first is that people don’t want to make the effort to do any calculation. They say “yeah, whatever,” and simply pick one of the three offered percentages (often the middle one).
The second explanation is that customers think the defaults establish the social norm. Many people want to give the right amount, fitting with what other people typically do; the defaults tell them what that is. The third explanation is that the amount imposes a kind of psychological pressure. Customers might think that if they go below the suggested percentages, they are expressing dissatisfaction with the ride.
Haggag and Paci suggest that all of these explanations are plausible, but their data don’t allow the authors to make a clear choice among them. Each of the explanations probably plays a role.
Unfortunately for drivers, the relatively high defaults produce a 1.7 percentage-point increase in the probability of a tip of zero. That backlash effect isn’t huge, and drivers are still way ahead on balance. (The 10 percent average increase includes the zero tips.) But tips of zero aren’t exactly welcome. It is reasonable to speculate that if the defaults were increased, the no-tip rides would increase as well.
These findings have implications well beyond the area of tipping behavior. We are in the midst of an explosion of research on the power of default rules. In the area of savings, defaults are crucial.
When workers are asked whether they want to enroll in 401(k) plans, the number of employees who do so is relatively low, certainly in the early years of employment. Recently, a number of companies have changed the default to automatic enrollment. The results are clear: Many more workers end up enrolled and are thus put in a position to retire with far greater savings.
Or consider this question, now attracting a lot of attention in the U.S. and Europe: How much privacy does any of us have online? The answer may well be a function of the default rule. Suppose that a website says your personal information won’t be shared with anyone unless you click on a button to allow it. Now suppose instead that the same website says your personal information will be shared with the entire world unless you click on a button to forbid it. Will the results be the same?
Far from it. If people are given a privacy-protective default rule and asked whether they want to shift and opt into information sharing, a lot of them will respond with some version of “No, you’ve got to be kidding.” In that case, their information won’t be shared. If, by contrast, people are given an information-sharing default rule and asked whether they want to opt into privacy protection, a lot of them will just stick with the default, especially if they have to think a little bit and read something complicated in order to switch. In that case, their information will be shared.
The broadest lesson is that for better or for worse, default rules and settings have a great deal of power. Businesses and governments need to think hard about them. Have any doubts? Just ask New York’s cab drivers.
(Cass R. Sunstein, the Robert Walmsley University Professor at Harvard Law School, is a Bloomberg View columnist. He is the former administrator of the White House Office of Information and Regulatory Affairs, the co-author of “Nudge” and author of “Simpler: The Future of Government,” just published by Simon and Shuster. The opinions expressed are his own.)
To contact the writer of this article: Cass R. Sunstein at firstname.lastname@example.org
To contact the editor responsible for this article: Katy Roberts at email@example.com