Saturday, July 27, 2013

EconOMNOMNOM: voting?

Welcome to EconOMNOMNOM (Econ Food for Thought!).  This is a place for silly ideas that I do feel like thinking about but don't feel like writing about. Just some stuff to munch on.

Today's EconOMNOMNOM:  What would happen if in presidential elections, not showing up meant you automatically voted for whatever party you were registered with?

Look, I'm not even going to spend time figuring out how to phrase that elegantly.  The whole point is for me not to have to do any real work here.  (Of course I welcome your comments and/or blog posts in response!)

It would also be great if I could make some money off of all this non-work.  Therefore I will be monetizing my EconOMNOMNOM posts with ads for delicious edibles so you can satisfy the OMNOMNOM cravings they invariably trigger.  This issue of EconOMNOMNOM is brought to you by Tony's Country Style Ham and Cheese Biscuitwiches.  This mouth-watering product really sells itself:
"Silver Lake and Michael Dell are asking that along with the increased offer, a special committee conducting Dell's sale agree to revise the merger agreement so absentee votes at the shareholder meeting aren't counted against the takeover consortium. It is reported that about 20% of Dell's shareholders were absent the July 18 vote, creating a roadblock for the private equity buyers."
...Not surprisingly, turnout really matters in elections where you vote "no" if you don't show up.

(It's good it sells itself because I certainly am not going to do any additional work, if you didn't get that already.  By the way, Tony, I learned why you find ham so mouth-watering: the ham you know and love is actually a ham-and-water-product!)

Sunday, July 21, 2013

Monterey Bay, and Penguin Cyclone!

Ryan visits the Monterey Bay aquarium of awesomeness!  When I was at Monterey Bay in 2004, they were the only aquarium to have (temporarily) a great white shark.  I also remember the jellyfish and Ryan's sardine cyclone.  Also they have penguins!

The Monterey Bay aquarium is probably the best aquarium I've been to.  The only reason for the "probably" is that I have also been to this aquarium in New Zealand.  (Yes, that is a conveyor belt that takes you through the shark tank.)

Anyway, speaking of herding and penguins, consider these emperor penguins, which huddle for a very different reason than the sardines, namely that it's cold in Antarctica:






Coordination Success!
What interests me is how the penguins coordinate their efforts.  The issue is that penguins on the outside of the huddle are exposed to the cold, so the penguins need some way of taking turns.  Importantly, the equilibrium must be robust to self-interested penguins who just want to stay in the center all the time.

According to the narrator at the 24 second mark, "Speeding up the action reveals an unparalleled display of sharing, as each penguin works its way toward the warmer center."

But that's not what's happening at all, is it?  Look at the 33-second mark.  Penguins aren't working their way into the center so much as the center is working its way to the penguins.  Here's how it happens:

  • A dissatisfied penguin on the outside can't hope to work its way into the center -- the huddle is very dense to conserve heat.
  • However, a penguin on the outside can easily move to another point on the outside.
  • In isolation that accomplishes nothing, but if all the dissatisfied penguins move to the same place on the outside, that becomes the new inside.

This is an equilibrium, and one that requires no cooperation, no "unparalleled display of sharing."  If penguins on the left edge of the huddle are constantly moving to the right edge, it is optimal for any individual penguin on the left edge to move to the right, because other penguins will soon surround him.  In this manner, the center of the huddle slowly shifts to the right, and every penguin gets its time in the center.

This is such a neat solution to the problem.  Penguins never have to fight for a good spot, which would waste precious energy in zero sum conflict, which is just plain infeasible given that the penguins have to play this game in the bitter cold for two whole months with No Food.  And the solution requires no benevolent penguin ever saying, "Hey, I've had my turn, why don't you take my spot now."  Which is good, because evolution has a hard time producing such creatures.

As a bonus, notice the penguin cyclone at the 18-second mark.  That's how you absorb a bunch of additional penguins into your group while staying in the center yourself!


Coordination failure?
Herding is always partly about copying your neighbors.  But in the penguin herd, copying your neighbors serves a particularly vital role, because coordination is necessary for this reshuffling to work.

Yet at first glance, there's a lot of disorder in the video above, with multiple groups of penguins sometimes moving about in seemingly random ways.  Does that disconfirm my model?  Not really.  I would say it's just plain hard for the whole group of penguins to perfectly coordinate their behaviors.  If you look closely, you will at least see that nearby penguins are generally doing the same thing as each other.  Penguins on the move are copying each other, walking around to attach themselves to the same side of the huddle.  The fact that they're copying each other ensures it is locally optimal to do so.

Coordination success!

Well, there is one thing that can help the penguins coordinate perfectly...a cold winter wind!





When a strong wind is whipping at the penguins, naturally the ones on the exposed side move to the sheltered side.  Perfect coordination when it matters most!

Added: Cold wind sucks, but it is theoretically possible for the penguins to be better off when the wind is blowing, if its value as a coordination device exceeds the damage done by the coldness of the wind.

Free bio research idea: It should be easy to confirm if penguins use the wind as a coordination device.  Just monitor their behavior, and plot some measure of coordination against wind speed.  Is there an optimal nonzero windspeed that gives them most of the benefits of coordination without too much cost?