March 2006
Sun Mon Tue Wed Thu Fri Sat
      1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31  

Authors' Committee


Matt Blackwell (Gov)


Martin Andersen (HealthPol)
Kevin Bartz (Stats)
Deirdre Bloome (Social Policy)
John Graves (HealthPol)
Rich Nielsen (Gov)
Maya Sen (Gov)
Gary King (Gov)

Weekly Research Workshop Sponsors

Alberto Abadie, Lee Fleming, Adam Glynn, Guido Imbens, Gary King, Arthur Spirling, Jamie Robins, Don Rubin, Chris Winship

Weekly Workshop Schedule

Recent Comments

Recent Entries



SMR Blog
Brad DeLong
Cognitive Daily
Complexity & Social Networks
Developing Intelligence
The Education Wonks
Empirical Legal Studies
Free Exchange
Health Care Economist
Junk Charts
Language Log
Law & Econ Prof Blog
Machine Learning (Theory)
Marginal Revolution
Mixing Memory
Mystery Pollster
New Economist
Political Arithmetik
Political Science Methods
Pure Pedantry
Science & Law Blog
Simon Jackman
Social Science++
Statistical modeling, causal inference, and social science



Powered by
Movable Type 4.24-en

« March 3, 2006 | Main | March 7, 2006 »

6 March 2006

An Unintended Potential Consequence of School Desegregation

Felix Elwert

One goal of school desegregation is to promote racial understanding by fostering interracial contact. In an article in the American Journal of Sociology (1998, Vol. 103[5]), Scott Feld and William Carter develop a simple combinatorial argument about a surprising potential consequence of school desegregation.

They argue that under certain (not so outlandish) circumstances, school desegregation may actually decrease rather than increase opportunities for interracial contact.

Here is their argument by way of a stylized example. Suppose there are four schools, one with capacity C1=400, and three schools with capacities C2=C3=C4=200 students. Under segregation, all 100 black students in the district attend the big school. The 900 other students are white. Assuming that students only interact with students in their own school, there are thus 300*100=30,000 possible interracial, intra-school ties. Now desegregate such that the percentage of black students is the same in all four schools. Then there are 360*40 potential interracial, intra-school friendships in the big school, and 180*20 potential interracial, intra-school friendships in each of the three small schools. Hence, the total number of potential interracial friendships post-desegregation is 25,200, as compared to 30,000 pre-desegregation.

Whether this decrease in potential ties will actually result in a decrease in realized ties is an empirical question, dependent on factors spelled out in the article. Feld and Carter go on to show that this particularly instance is an example of the so-called Class Size Paradox, known from various applications in sociology.

Posted by Felix Elwert at 6:00 AM