Title: An empirical investigation of the relationship between the laws 
of Zipf and Gibrat

Abstract: The talk will present empirical evidence against currently 
popular explanations of Zipf's law as the asymptotic outcome of appropriately
modified versions of Gibrat's law. The laws are studied in the context 
of US city and firm growth and the empirical analysis is based on a new form 
of the variance ratio test modified for use with panel data. This motivates
work in progress that aims at a different type of explanation for 
Zipf's law that is inspired from biological scaling models.