Before I get to the article I like, I’ll talk a little about the context of why I like it.

Economics and psychology together, i.e. behavioural economics, has long known that post-facto “rewards” for behaviour is usually only effective if the person knows in advance what the reward is going to be. So, if you set a goal, and the person values it, they will engage in the behaviour required to “win” or “earn” the reward. Gamification only works if the person knows the rules and has some say in the reward, i.e. it isn’t random chance.

Yet around the world, “tipping” doesn’t follow that pattern. It is an unknown reward provided after the transaction (i.e. the meal, for the restaurant world), and is supposed to reflect the customer’s view of how well they were served. Better service, better tip. Poorer service, poorer (or no) tip. Yet people rarely deviate from the norms — they often will pay 10% or 15% or 20% all of the time, by their personal comfort levels, for the wide “middle” ground for the level of service. And in some cases, they do it not because they think it is the “approximate” value of the service, but simply because it is the recommended norm for “good” (often average) service.

So some people balk. They get poor service, they stiff the server. Maybe the food was cold, maybe it was late, maybe it was the wrong order. Interestingly, speaking to experienced servers, one of those events (late) isn’t even their fault (if it is cold, it’s because it was sitting somewhere waiting to come out; if it is wrong, it is because the server didn’t check the order before bringing it out, according to those experienced in the industry). They also don’t control staffing levels…so if it is normally a 6:1 ratio for tables: servers, and because two servers are away or they haven’t been able to staff the spots, and suddenly it is 9 or 10:1, that’s not really the server’s fault either. Yet it is their tips that will suffer. A variable reward, random chance in some cases, having little to do with their performance. Most people don’t have any idea if their server was good, they just know if the overall experience was satisfying. Maybe they were sitting next to a cold window or a fan, maybe they’re sensitive to bright lights, maybe they’re fighting with their spouse, and it affects their view of the meal. Or they wanted more veggie options than the restaurant has on the menu.

I have some problems with tipping as a norm, I confess, for four reasons.

First, in Canada at least, with the automated payment machines, when you type in your amount of say “15%” for it to calculate for you, it does so on top of the tax. Why would I pay a server “extra” for the restaurant collecting tax? That makes no sense. Is it a big amount? No, of course not. But if the harmonized tax rate is 15%, and you assume no alcohol to keep it simple, then your 15% tax adds 2.25% to the tip. So they get 17.25% instead of 15.

Second, I don’t like flat-rate commissions with no top end. I don’t really like it with real estate, I don’t like it with any commissioned sales, really. If I go out for dinner with my wife and son, and we go to a simple restaurant, and the bill for dinner is $60, then the “standard” of 15% is $9. If it is a nicer restaurant, the bill might be $100, and the standard tip would be $15. Did the server who served me the more expensive food do any more work than the server who served me cheaper food? Why am I paying him/her 2/3 more in tip? Perhaps their costs are higher (better clothes and shoes, etc.), but a 2/3 increase in tip? People object on the same basis for paying commissions on house transactions — is there more work involved if the house is $700,000 than $200,000? Should the agent get an extra $12,500 for helping with a more expensive home? Taxes might be proportional, but why are the commissions, separate from the perverse incentives that are created?

Third, the hourly wage for the server varies dramatically across a shift, and generally for no real reason other than the equivalent of “piece work”. When they’re busy, their tips are high, and their hourly wage soars; when it’s dead, they make nothing, even though they’re still at work. Piece-work in factories is generally viewed as highly exploitative, particularly when the “pieces” are not all made by one person i.e. an attribution problem. Similar to the comments above, the server doesn’t buy, prepare, or cook the food, nor clean up afterwards. Yet they are the “face” of the service, so they get tipped accordingly (good food, good tip; bad food, bad tip). But let’s ignore base wages for a second and look at a restaurant shift from 5 to 10 p.m. at night. From 5-6, the server might have 3 tables. From 6-7, perhaps that goes up to 6 tables, 7-8 goes back to 4, 8-9 is 3, 9-10 is 1. If we assume all the tables were tables of four with bills of $80, and they all tipped 15%, then the server would get:

• 5-6, 3 tables, $36 in tips;
• 6-7, 6 tables, $72 in tips;
• 7-8, 4 tables, $48 in tips;
• 8-9, 3 tables, $36 in tips;
• 9-10, 1 table, $12 in tips.

Yes, I’m exaggerating slightly, but some servers have been known to be able to handle 6, 8, 10 tables of 4 in a restaurant with a relatively static menu. In the busiest hour, the wage is $72 / hour. Really? We’re paying someone $70 per HOUR to deliver food to a table? For the night, though, they’re clearing only $204. Which, while not chickenfeed, reduces down to $40 per hour. Before tax, or any tip-sharing that goes on. Most servers I know have said even without tip sharing, they always gave some money to the bus people to incentivize them to clear their tables quickly and get the next group in. Their wage per hour though drops as low as $12 and goes as high as $72, totally based on foot-traffic, not their performance. It is exploitative, stressful, and chaotic for steady income. In the bars that serve food and a lot of drinks, the transaction totals are smaller, but the servers frequently make more. Partly because the “business” is steadier than the pure food totals.

Finally, though, we come to the article. I’ve often felt that I would prefer simply for the tip not so much to be “included” as just that the servers were paid a decent hourly wage. No tipping, or if still done, limited to something like 5%. A token amount to maintain some incentive I suppose. But the article belies all that, because it goes through a bunch of existing US statistics from the Department of Labor to show the reality of the service industry:

Tipping, while practiced around the world, assumes a unique role in America, one to which most diners are obliged, because the United States is one of the only countries that allows businesses to offload the burden of paying workers a fair wage to their customers. And though construed as a fair way to encourage hospitality and reward good service, tipping’s roots are in racialized exploitation, while recent data shows that it continues to be, at its core, racist, sexist, and degrading.

5 Ways Tipping Hurts Restaurant Servers and Customers

It is exploitative as it creates power plays between employers who control the opportunities without paying a living wage and the employees who earn the tips, but who are left vulnerable to mistreatment and abuse for back wages and pooled tipping managed by employers; it reflects and amplifies racial inequality and profiling (white servers are tipped more, whites get better service than blacks); and it fails to prevent and thus supports sexual harassment from customers, as the server is financially penalized if they push back.

I guess, in some ways, I just like the article as it unpacks the reality that it sucks for everyone, except maybe the employer.