help me scam my students

[Michael McIntyre]

Here's the scam.  For a course called "States, Markets, and Societies" I plan to walk in on the first day, have each student roll a pair of dice, take her/his name, and put the name and roll of the dice up on the board.  Then, with some suitably vague reference to Howard Becker's forty-year-old study of students and grades at the University of Kansas, I'll announce that since it's clear that grades are an impediment to learning, I've decided to get around the grading problem by assigning grades randomly.  Then I'll reveal the grading scale.  A roll of 12 is an A; a roll of 2 is an A-, 11 is a B+, 3 is a B, and so on down to 7, an F.  Barring some very strange throws of the dice, then, F will be among the most common grades.  This, I predict, will piss my students off.

After letting them vent for awhile, I'll tell them that the only way they are going to get me to change my policy is to convince everyone in the room that an injustice has been done.  "Everyone" will include the people lucky enough to roll for high grades.  We'll see if they have sufficient ingenuity to demonstrate that random, arbitrary distribution of a valued good is unjust.

At that point (maybe the following class, depending on how time goes), I'll introduce them to the notion of a birth lottery - the random and arbitrary assignment by birth of one's valued goods at least for the first couple of decades of life (and, substantially, for much longer than that).  Having just conclusively proved to their own satisfaction that this arrangement is unjust, the class, centered on the theme of inequality on a global scale, can proceed.

So how can you help?  Well, I'm looking for a little technical support here.  To make the second part work, it would be good to have data on the world distribution of income fine-grained enough to allow me to interpoate per capita incomes at specified percentile levels (17, 31, 44, 56, 67, 75, 83, 89, 94, and 97 to be exact).  If that data is out there, I haven't seen it.  I'm willing to go with relatively unrefined data (non-PPP adjusted, based on median national incomes, etc.) as long as I can use it to make a dirty ball-park interpolation.

Any ideas where I can get data like this?

Michael McIntyre

PS - If you want to know the reading list:
Robert Bates, Prosperity and Violence
Gianfranco Poggi, The Development of the Modern State
Charles Tilly, Coercion, Capital, and European States
Ellen Meiksins Wood, The Origin of Capitalism
Kenneth Pomeranz, The Great Divergence
Mike Davis, Late Victorian Holocausts