Most research on residential mobility has documented a clear pattern: Resid
ential and school moves are associated with poor academic performance. Expl
anations for this relationship, however remain speculative. Some researcher
s argue that moving affects so- cial relationships that are important to ac
ademic achievement. But the association between moving and school performan
ce may be spurious: the negative correlation may be a function of other cha
racteristics of people who move often. We offer several conceptual and anal
ytical refinements to these ideas, allowing us to produce more precise test
s than past researchers. Using longitudinal data, we find that differences
in achievement between movers and nonmovers are partially a result of decli
nes in social relationships experienced by students who move. Most of the n
egative effect of moving, however, is dice to preexisting differences betwe
en the two groups.