Category: Food and Drink

Eric Crampton emails me:

Why don't we see more of this?  I went to the only Ethiopian restaurant in New Zealand last night.  It runs one day a week – Mondays – in a Burmese restaurant that otherwise was closed on Mondays.
 
http://offsettingbehaviour.blogspot.com/2010/02/one-day-week-restaurant.html
 
I can understand that this kind of arrangements would have risks for the host restaurant.  Ideally, you'd want it from a non-competing cuisine style.  But this is the first instance of it I've ever seen.  Have I just not been paying attention?  The story from the Dominion Post on how the place opened is very nice.  The woman running the Burmese restaurant was an immigrant from Burma who later started volunteering with an NGO that helped new migrants acclimatize.  She met a guy there who wanted to open an Ethiopean restaurant but had no capital; her restaurant was closed Mondays.
 
The other 6 days a week the Ethiopean restauranteur drives a cab. 

I have an arbitrary aesthetic preference for the heavier, larger milk containers and now I am wondering if there is an efficiency argument for them as well.  A few times lately I've bought the smaller containers and — it seems to me — they turn out to be stale and sour pretty often.  That's hardly rigorous data, but I have been pondering theory.  Why might the larger containers be fresher?  I can think of a few reasons:

1. If you buy three small containers instead of one large container, you might end up consuming the oldest container last, leading to a stale result.  We are all sloppy in checking expiration dates.

2. The large containers are expected to sit around your fridge longer.  So when they are put out on the shelf by the grocery store, there is a greater margin of error built in, vis-a-vis their freshness.

3. When shoppers choose the larger cartons, they find it harder to paw through the selection and take home the freshest, thus leaving the stalest for other buyers and increasing the variance of freshness/staleness in the overall supply.  You are less likely to end up with "the leftovers."

4. Whether you buy large or small containers, there is "container overlap" in your refrigerator.  With the large containers, you can (and should!) sample the quality of your next container before finishing your previous container.  In doing so, you learn about the viability of your milk supply for days to come and you can react accordingly if the container-to-come is stale.  You won't and shouldn't open three smaller containers to get a comparable preview of the milk-to-come across such a long time period.

Have I missed any arguments?

How many of these arguments, if any, imply that married people are happier than single people with serial relationships?

Leigh Caldwell offers an analysis.  Here is one bit:

…why is it losing so much money when demand is so high? The 48-seat restaurant has a six-month season with about 8,000 covers a year. It receives 300,000 applications for those seats [though this article says a million and this one two million], selling out the whole year's reservations on the same day that bookings open for the season. Why wouldn't they bump up the price from 230 to 330 euros, to simultaneously manage demand and eliminate the losses? Price elasticity can't be that high.

My hypothesis is that the restaurant was never intended to turn a profit, but rather it was a loss leader for book sales, endorsements, lecture fees, TV contracts, cookware lines, and so on for Ferran Adria.  Even if higher prices could bring in a twenty percent rate of profit, it wouldn't — at this point — be worth keeping the place up and running.  Adria already has a reputation as the world's greatest chef, running the world's greatest restaurant.  It's best to quit while ahead and branch out into food-related money-making ventures.

The low prices make going a hard-to-obtain event, open up the restaurant to more people than just the very wealthy, and maximize the publicity value of Adria's name.

He won't and can't stop cooking forever, but cooking six months a year is probably not an optimum for him at this point.  The real profit and loss calculation for El Bulli has to include the shadow price of his labor as an important variable.

Bryan Caplan raises an excellent puzzle, following his visit to the supermarket (I can confirm similar observations, even earlier in the day):

They were out of milk and bread, but there was still plenty of cheese and chocolate.  That was easily explained – people knew they could shop again in a few days, so they only needed to stock up on staples.  But the more I looked around, the more puzzled I was.

Here's what I noticed: For any given type of product, the most popular brand always sold out first.  There were no Eggo waffles, but plenty of Wegmans brand waffles.  All the national brands of hot dogs and sausages were gone, but there were plenty of obscure sausages still on the shelves.  If you broadened the categories, the pattern remained.  In produce, all the bananas were gone, but there were still plenty of apples.

You might say, "What's the puzzle?  Of course the most popular stuff sells out first."  But that's a feeble explanation.  After all, if X is ten times more popular than Y, then you'd expect stores to simply carry ten times as much X as Y.  Why would X sell out faster in a blizzard if stores have already taken its greater popularity into account?

I see it like this.  When visits to the store are in "normal times," the store weighs "mass goods" vs. "niche goods" when stocking the shelves.  Some niche goods will be given shelf space because they get some minority of customers into the store in the first place.  (I go to Whole Foods for Spelt Flakes, two or three key cheeses, and my favorite dark chocolate.  While I'm there, I end up paying the higher prices for their milk and green peppers, neither of which is better for me than competing products at Shoppers Food Warehouse.)  If the store stocks too few niche goods, it attracts too few niche-oriented customers and loses money on milk, meat, etc.  

Note also that the niche buyers tend to hold stocks of their favored goods, so if they have a sudden, unexpected supermarket trip, they don't necessarily need to buy more.

Enter the snowstorm.  People are forced to visit the store, more or less.  The previous calculations of mass vs. niche goods are no longer appropriate for the new emergency.  The store, temporarily, would prefer to have more mass goods and fewer niche goods.  The niche goods served the function of "motivating 47 visits a year rather than 23" but in the new short run they are nearly useless for this purpose.  We will see too many of them left on the shelves.  

Read Bryan's comments section as well.  I pondered the "rate of restocking" answers, but slow turnover and slow inventories imply a higher profit margin (if the product can compete for shelf space) and I believe for this to work it has to invoke customer heterogeneity in some manner, as I have done above.

Addendum: Imagine the same problem in the context of a book store.  There is an emergency edict which requires everyone to read three newly purchased books over the next week.  Borders will be swamped, They will end up short on bestsellers, not their niche books.

Here is the abstract, I look forward to reading the paper:

It is argued that drug consumption, most commonly alcohol drinking, can be a technology to give up some control over one’s actions and words. It can be employed by trustworthy players to reveal their type. Similarly alcohol can function as a “social lubricant” and faciliate type revelation in conversations. It is shown that both separating and pooling equilibria can exist; as opposed to the classic results in the literature, a pooling equilibrium is still informative. Drugs which allow a gradual loss of control by appropriate doses and for which moderate consumption is not addictive are particularly suitable because the consumption can be easily observed and reciprocated and is unlikely to occur out of the social context. There is a trade-off between the efficiency gains due to the signaling effect and the loss of productivity associated with intoxication. Long run evolutionary equilibria of the type distribution are considered. If coordination on an exclusive technology is efficient, social norms or laws can raise efficiency by legalizing only one drug.

I thank Brian Dailey for the pointer.