To Tip or Not to Tip

EconTalk Extra
by Amy Willis
Anthony Gill on Tipping... Tim Harford on Fifty Invention...

kids tip.jpg Is tipping a relic of the past that has outlived its usefulness? Should restaurant servers work for tips or a living wage? This week, EconTalk host Russ Roberts welcomed back political scientist Anthony Gill to discuss his recent piece on teaching about tipping.

Now we'd like to hear more from you. Use the prompts here to share your reaction to this week's episode, or to spark your own conversation offline. Feel free to post your own questions here, too. We'd love to converse with you.

1. Did your parents teach you to tip for service? If so, what was their rationale, and was there a guideline for how much to tip?

2. What are the norms regarding tipping in your area? For what service(s) do you tip, and how much? Have you ever found yourself in a situation where you "got it wrong?" What happened, and how did it influence your future behavior?

3. How does tipping allow for voluntary price discrimination, according to Gill, and on what grounds does Roberts disagree? With whom do you agree, and why?

4. Why would people tip in a place they never expect to visit again? (And why does Roberts jest that "only an economist would think this is a puzzle?")

5. Special bonus question for all you econ teachers out there...Gill discusses tipping as a teaching tool in his political economy class. It's a real-world example he uses to help his students understand concepts such as the principal-agent problem and price discrimination. What are your favorite real world illustrations of economics concepts, and how do you use them?

Comments and Sharing

CATEGORIES: Extras (194)

TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (14 to date)
Shayne Cook writes:

#2. ... situation of "getting it wrong" (sort of)

I spent about a month and a half in Europe this Summer - Ireland (both), Scotland, England, France and Switzerland. The trip was for pleasure, not business, and some of the visit was part of a guided tour, part was on my own.

More than once, both guided and on my own, I was advised - "reminded", really - that Americans are noted for "over-tipping". And I was advised and admonished to avoid that.

Folks in the UK and CH in particular consider American over-tipping a typical indication of American arrogance and "superiority" - "showing off", both financially and otherwise. As such, American over-tipping is insulting.

I was advised, more than once, that about 5% is closer to "recommended norms" there, than 15% - with about 10% being a sort of "threshold of obvious just-showing-off".

Different places/situations and different "norms", of course.

But the more interesting dynamic, in context with Russ' and Gill's discussion here, is: What portion of American tipping, and even tipping "norms", is "compensation" and what portion is just "showing off", a signaling of "financial superiority" - even within the U.S.?

Robert Herreid writes:

The worse, the better?
Fine dining with tipping is in a death spiral wherever laws make owners and kitchen staff the adversaries of waitstaff (everywhere in the US). Owners have very little say over the conduct of waitstaff who see themselves realistically as the employees of customers. At at a high end place, the portion of waitstaff income coming from owner is trivial. Some waitstaff refuse to do anything that will not be a reason for a tip. As minimum wage gets higher and the wage differential between kitchen staff and waitstaff gets bigger, restaurants will close.

Jerm writes:

2) I am from Hawaii, where there's a strong tradition of non-cash tipping. When I was younger, I was taught to pack gifts for the hotel maids in my luggage (normally, boxes of chocolate-covered macadamia nuts). I know tons of people who do that, even if they also leave a cash tip.

Once I tipped a woman at Waffle House an entire can of macadamia nuts. She was not expecting that. In retrospect, I over-tipped.

On a trip a couple of years ago, I packed some "only-sold-in-Hawaii" Bath and Body Works hand sanitizers. They're small. It's something that they couldn't get outside of Hawaii. And I would include that with my regular tip, even at restaurants.

At one place, I was a repeat customer. And someone said "Hey, are you that guy from Hawaii?" I was. "Are you carrying macadamia nuts?" I wasn't.

So from now on, I always travel with a spare can. Just in case. People have expectations, after all.

Jerm writes:

5) When discussing the principal-agent problem, I like to give the example of taxis. This is because taxis in Hawaii are often rented by the driver (for 24 hours), and he then has the incentive to work himself to exhaustion over the next day.

Makes you wonder why restaurants whose waiters make hundreds of dollars in tips per night don't just treat their waitstaff as independent contractors and CHARGE them for the opportunity to work.

Also, with the popularity of Uber, it's a perfect example of the backward-bending supply curve. There's one or two Uber drivers in each class, and they've always said that they drive only enough to make a certain amount of money for the night and then STOP. The surge pricing doesn't get them to get off the couch or to drive more.

Anthony Gill writes:

Jerm has an excellent question above as to why wait staff are just contracted out. (This also relates a bit to Robert Herreid point.)

First, there was at least one moment in time where wait staff were contracted out as such. Kerry Segrave writes in his book Tipping: An American Social History of Gratuities, that during the early 1930s at "restaurants along the Champs Elysees, waiters paid ten to 12 francs a day for the privilege of working" (p. 49). Segrave also noted that coat checks at various East Coast hotels were contracted out as well, and not run by the hotel. Tipping would be the main source of income that would be shared by the employees and the owner of the coat check franchise.

While seemingly an ingenious idea, I can see how this would relate to the problem Herreid writes about above. If wait staff were solely paid on tips, they would have very little (if any) incentive to listen to the owner or kitchen staff. If you pay the wait staff a wage, you then can exercise some control over them. This reminds me of Michael Munger's podcast on the nature of the firm and transaction costs. If the wait staff was only paid by tips, then they would have more freedom when they wanted to work. There obviously would be some key times that would be ideal, but I don't think many wait staff would want to work a slow shift on a late Tuesday evening. By having some of the compensation paid partly in wages, the owner would have more ability to "punish" (withhold compensation) wait staff that didn't want to work particular shifts. It probably makes shift planning better. Also, it gives some leverage to the owner for incentivizing the wait staff to work more effectively with the back of house. Of course, the owner could refuse to let recalcitrant wait staff to work at the restaurant at all, but being able to reward/punish on another dimension other than table service may help smooth out some of the other problems that may arise.

Anthony Gill writes:

Allow me to give a stab at #3: How does tipping allow for voluntary price discrimination, according to Gill, and on what grounds does Roberts disagree? With whom do you agree, and why?

I obviously agree with Gill. Here's why.

I view price discrimination as the ability of a seller to charge different prices to different consumers based upon the different reserve prices of the consumer, which helps shift the gains-from-trade in the direction of the seller.

Assume you have a seller who has a reserve sell price of a cup of coffee equal to $1 (i.e., the seller won't go below that price). And then you have two consumers, one who values the coffee at $1.25 and the other at $4.00. The seller would like to charge the first consumer $1.25 and the second consumer $4.00. The trick is teasing out who is who and prevent the second consumer from lowering his revealed reserve price with knowledge of what the first consumer is buying at. Obviously, some degree of monopoly power will enable this, but with asymmetric information it is difficult for the seller to tease out what people will pay. Sellers might try different tricks such as the use of coupons, special sales at different times of the day, etc.

In the case of tipping, I argue that restaurant owners will want to keep their reserve price as low as they can to make sure they can attract customers, since empty seats are dangerous for a restaurant -- tables not used, staff have to hang around for a potential "rush," and it may signal that the restaurant is not good. Richard McKenzie and Steven Landsburg have pointed out that restaurants will vary their profit margins on various products (dessert, wine) based upon how much time somebody might consume at a table.

If the owner can shift some of the cost of labor to the customer on a voluntary basis (i.e., tipping), they are able to keep their overall prices low and allow the diner to determine what they will pay for service. Some people will be generous tippers because they value service highly. Others might not tip as much (e.g. 10%), BUT they still come into the restaurant because prices are fairly low. Wait staff might not like this person compared to high tippers, but at least they are filling a seat (and purchasing food) that might otherwise go empty (or unpurchased), which is important to the owner.

I see a strong cultural norm of tipping that allows some variation in range to allow the customer to voluntarily "price discriminate" based upon how they value service. If they value service highly and it was good service, the norm of treating people well who treat you well would mean dropping a 25% tip. This makes both the wait staff happy and the owner. (Some of these folks also get special food benefits because they are known to be good customers and everybody wants them back.) I just find this ability for me to set the price and share the gains-from-trade to be really cool.

I should note that if a restaurant went to a "no gratuities" model, and set the service portion of a meal at a fixed price, there is a possibility that this would be well below the reserve price of generous tippers who might otherwise leave more at the table. In this instance, a "no gratuities" model benefits these "big spenders" at the expense of the wait staff. To the extent that wealthy people tip more (and this is just an assumption), wealthy people would benefit at the expense of the poor. Some poor folks who may be not so good at tipping (or individuals who are trying to save a few bucks but still want to go out for dinner) might be priced out of the market. (You can see this graphically in the SSRN version of the paper.)

I do cite an article by Lott and Roberts about their skepticism on some pricing schemes that are attributed to price discrimination (and that it might just be unobserved costs). There is much to say on this. But I do think there is a bit of a "dance" that goes on between the seller and buyer in trying to figure out how to get the most of the gains-from-trade. We do this at auto dealership and in the real estate market, why not at a different (and scaled down) level in other businesses? To some extent, as well, I think we over-emphasize the "monopoly power" component of price discrimination; entrepreneurs can be pretty creative in figuring out how to shift that consumer surplus in their direction.

Margaret H. writes:

The principle-agent problem does not work itself out quite so neatly in a real life restaurant scenario. In my own experience, I can recall a server in my restaurant who might have been the “perfect” waiter. He was able to read people like a book. He knew who wanted to be chatted up, who wanted to be left alone, who wanted to be flirted with, who wanted to get in-and-out in a hurry. This excellent waiter was so adept at reading people, he could read what their salary was and what price range they would be likely to stay within for a dinner out. His wonderful and attentive service he reserved for the suits and wealthy older clientele; the tables that were going to, at best, order a soup and salad each were not worth his time, so why should waste his charms on them? As I see it, the only way the principle-agent problem could be solved by tipping is if the customer playing field were to be leveled. If only wealthy business men ate at our restaurant, I would have had the perfect waiter.

Robert Herreid writes:

Who is the "principal" in principal-agent analysis? Tipping does not prevent waitstaff from slacking off from doing what the owner wants done. In fact it makes it unlikely that waitstaff will willingly fold napkins, clean work stations or anything else that doesn't generate a tip (or likely to generate smaller tips as in working weekday lunch shifts).

The system now, with employers taxed for tips and tips not allowed to be shared (thanks to the government) works fine for the waitstaff at the expense of the owner and kitchen staff. It's not a win, win, win; but win, lose, lose. Every increase in tips is a negative for the owner. It means an increase in taxes, and therefore less money available to pay the kitchen staff.

Tipping is the largest part of a waitstaff's income. He who pays the piper calls the tune. Is the owner the principal, or is the customer?

Jerm writes:

I don't see why a (minimum) wage would make a server more likely to cooperate with the back of the house. The main incentive for those workers to cooperate is that they want to work again in the future. And an independent contractor can be dropped from the schedule far easier than a regular employee.

As for filling the slow nights, it would be simple. In Waikiki, it's not unusual to make about $100 in tips on a weeknight and $400 on a weekend night. So the restaurant could charge (their independent-contractor-servers) $20 to work on a Tuesday and $100 to work on a Saturday. Or you could use the slow nights to try out the new workers. Or you could bundle the days together so that a server has to do the slow days in order to get the big ticket days. A lot of these strategies exist already, but a contracting model would add another tool to the owner. And it's much more flexible.

If this is truly an example of the principal-agent problem, it seems odd that the classroom solution to this problem is so quickly thrown away.

Amy Willis writes:

Jerm I'd be curious to hear more about the non-cash tipping tradition. On the one hand, I would to get such unique gifts (well, except macs, but only because I'm allergic!). But on the other, in places outside Hawaii, does this "crowd out" real income? You note in your comment that stateside you include these non-traditional tips along with cash, but how does this work on the islands?

Mezza writes:

In Hawaii, it was customary to leave the trash collector a Christmas tip - typically a 12 pack of beer. I never could figure out how the garbage men fit all the beer in their truck cabs.

Frank writes:

I generally tip wait staff 18% as vase and up or down based on service.

At hotels. I don’t always, sometimes I forget. When I do I use the small notepad to write thank you and a smiley and leave the money next to it.


I do tend to overtip a little. I justify this, in that a 20-25% tip may genuinely make a person feel very good, even if it is just a few dollars. On small bills, such as breakfast, I may even tip more, rounding up to $25 on a $19 check. Also, at least in part, a dollar or $3 or $5 or even $10 does not mean as much to me as it does to that person.

Carsten writes:

I found it fascinating that two economists (economists who usually like to do all kinds of a/b experiments to figure out what work and what doesn't) talk about tipping for an hour and don't really talk about how things are in other countries. ok it was mentioned that tipping amounts differ across the world, but if that is the case then why not use that to figure out if tipping (or close to zero tipping) has an impact on service -- does service in Germany or France suck compared to the US for example? -- or on the topic of reducing empty seats by attracting more customers? -- again are there less empty tables in the US compared to again Germany or France.

I think one could very simply figure out if the benefits of tipping actually work, or have the impact they are desired to create.

I would actually say that in the US tipping has created a large amount of decent service staff, but that in countries where tipping is not the norm one finds much better service staff at the top end since those are actually professional servers or hotel staff rather than hired hands.

Steve Kelley writes:

One aspect of tipping I was hoping would be covered is the concept of a sliding scale based on the size of the ticket in bars and restaurants. When two servers are working equally hard, spending just as much time with my table, refilling my drink the same number of times, etc. I am challenged to understand the disparity in their tip based on the price of what I ordered or the prices established by the owner.

The waitstaff does not have capital at risk in the form of investment in the food and beverage I'm being served, they are not contributing to the higher overhead of finer restaurants, they are simply providing a service not dissimilar in function from the same service provided in restaurants of all price ranges where table service is provided. To base their agency compensation as a derivative of the bill (often including a substantial portion of which is often tax) defies all other forms of compensation theory.

Certainly it drives the best wait staff to gain and maintain employment at the finest restaurants, but I still view it somewhat punitive to workers in lower priced establishments and unjust enrichment for servers at higher end places. At the end of the day, my hope and expectation is for quality service at either. Don't we deserve that as patrons?

My personal habit is to tip a higher percentage at lower price restaurants if the service warrants - typically closer to 25% for bills up to $20 to $30. But for bills of $100, paying $15-$20 for basically the same activity (the same specialization in the division of labor) seems irrational. A tip of 10% at the second restaurant still provides a premium of roughly 30-50% over the original scenario, a pretty solid return for simply having gained employment at a higher priced restaurant.

For convention I am not comfortable tipping 10% on higher priced bills, but I do typically back out the tax (why tip on tax?), pay corkage of $5 or so per bottle of wine removing the cost per bottle from the tab, and then tip roughly 15-18% on the residual of bills over $100, which usually bring the total tip closer to 13-14% of the bill "all in."

Comments for this podcast episode have been closed
Return to top