Which City has the Most Dis-Honest Tea Drinkers?

According to an experiment by Honest Tea, it’s L.A.

The company has placed unattended racks of its cold bottled tea on street corners in a handful of cities. A sign asks people to pay $1 per bottle, a heavy discount already. Viewers then “watched people wrestle with their conscience.” Hidden cameras live-stream the action here.

So far, the citizens of Seattle are coming out as the most honest, with 97% of people paying. Atlanta, Boston, Dallas and Cincinnati are in second with 96%. L.A. is last with 87% — they were actually at 90% earlier today, but that fell as the day went on, and the temperature went up. Here are yesterday’s high temperatures for the handful of cities, in order of payment rates:

Seattle: 67° F

Atlanta: 95° F

Boston: 86° F

Dallas: 101° F

Cincinnati: 100° F

Washington DC: 94° F

Philadelphia: 92° F

San Francisco: 74° F

Chicago: 85° F

Miami: 92° F

New York: 95° F

Los Angeles: 92° F

By that measure, Cincinnati and Dallas, where the temperature was over 100° F, look even better than say, San Francisco, where it was a cool 74° F. Also, should we be less impressed with Seattle since it was only 67° F and raining, and people were probably less thirsty? It would be worth doing this again when we weren’t in the middle of a summer heat wave, to see if payment rates go up as temperatures go down.

UPDATE: It appears that by Wednesday morning, Chicago had vaulted into the lead with a 99% payment rate, while New York sunk below L.A. for last place, with an 86% payment rate.

Here’s the company’s promo video of the experiment:

[HT: Robert Reinheimer]


Joshua Northey

How does it effect the honesty rating if someone steals the money box but doesn't take the tea?

I would also think the conspicuousness or non-conspicuousness of the cameras, the openness and traffic at the location, and income of passersby would have very large effects. Larger then the heat.

Chris Sampson

As temperature rises, individuals' willingness to pay should go up... surely making them more likely to pay, not less. The people of Seattle must just be wonderfully nice.


I see two reasons why a rising temp should cause more non-payment: Firstly, that otherwise honest people who happen to not have any cash at the moment may be driven to desperation by the heat and compromise their morals in exchange for a refreshing drink. Secondly, that the heat saps one’s mental energy making harder to control one’s selfish impulses.

I also think there are a lot more factors, like placement of the stands, that need to be accounted for.


Just from watching the video feed it's obvious that there are some employees of some kind dressed in t-shirts and shorts (no logos or uniforms) restocking the shelves with cold tea from the coolers. I even saw some peopel try to give them money for a tea. There also seemed to be some confusion about exactly what was going on, and the people working the stand looked like they might've been explaining things to confused 'customers'.

It would be interesting to see what kind of effect a more automated system would have on honesty.


What? People are dishonest in silicone valley?

But seriously, who carries cash? Maybe if they'd had a card swipe (Square?) it would have been more reasonable.

Paul M.

In LA Honest Tea is given out at many of the events I attend so I'm inclined to believe the stuff is free if it's out in the open.

Brent (in Seattle)

Although it seems to make sense at first thought, on second thought it's hard to understand as an economist why you hypothesize that people might pay more as temperatures fell. Lower temperatures reduces the opportunity cost of not having the tea. If you're near death from thirst in 100 degree temperatures, you'd pay, say, $1,000 for the tea; if it's 40 degrees and you just had a gallon of liquid you'd be willing to pay closer to zero (so you engage in the transaction for "free" stolen tea). Overall, given equal chances of being caught, it's not clear why temperature would matter at all...there is no income effect or substitution effect on your preference curve. If you prove to be empirically right in your hypothesis, then perhaps it tells us that a guilty conscience is a form of "payment." "Total payment" = Cash + guilt. As temperature go up, guilt goes down, cash stays the same (zero) encouraging engagement in the transaction. So people who will pay (high guilt cost) buy the same under all temperature conditions, yet those with low guilt costs are more apt to steal when the temperatures rise.

Interesting stuff. Reaction?



This would be true if it were a pay-what-you-want model, but the scenario is more like this: You're walking down the street with no cash on you, and see this stand. If it's 65 degrees out, you'd probably walk by without taking one. At 100 degrees, you would probably be more desperate and take one, even though you had no money.

caleb b


Elephant in the room: I want to see the demographics data when this is complete. Men vs Women. Observable races. People wearing professional clothing, people not. Then move the stand to richer and poorer areas of the city. Let the data speak for itself.


Other relevant factors other than the placement of the stands, traffic patterns and temperature:

- the numbers and movements of homeless people
- the relative frequency of, or familiarity people of the city have with, cash-based transactions, vs. swipe-and-go.


also in my local grocery store, honest tea is discounted to $1 per bottle at least once a month. it's not really a "heavy discount" unless you always believe the salesman that always say, "prices are low low low, you gotta jump on this now!"


Speaking of Honest Tea, has anyone else noticed the lie on their labels? (hint: graphical representations of statistical data)


'Honesty through paranoia' - Dante Hicks ;)


They did something similar in New Zealand - except that some people stole all the money out of the honesty box.

Marcus Kalka

Interesting study. Good marketing concept as well.

I'm now curious to know whether dishonest people prefer Coke or Pepsi.


I wonder how much the product name of Honest Tea affects the honesty of the customers. What if the brand was Generic Tea, or for the Econ nerds, Heteroskedastici Tea (the tea with more varying flavors with each gulp)?

My point is that someone who is about to "steal" a tea might think differently if the name of the product they are about to steal is called Honesty. If the same experiment was run with a different product name, it could be that the honesty results would be lower at a significant level.


I'm sure I read a study years back about about some profesional who got made redundant and as he couldn't find another job he spent his time setting up honesty boxes in offices with bagels and cream cheese, as this had been popular in his old workplace.

He left the boxes in the morning with a notice saying it was an honesty system and went back in the evenings to collect whatever money was left.

He found that the larger the organisation the less honest the people were.


67 and raining IS a heat wave in Seattle.