NATIONAL BUREAU OF ECONOMIC RESEARCH
NATIONAL BUREAU OF ECONOMIC RESEARCH

Rank-1/2: A Simple Way to Improve the OLS Estimation of Tail Exponents

Xavier Gabaix, Rustam Ibragimov

NBER Technical Working Paper No. 342
Issued in September 2007
NBER Program(s):   TWP

Despite the availability of more sophisticated methods, a popular way to estimate a Pareto exponent is still to run an OLS regression: log(Rank)=a-b log(Size), and take b as an estimate of the Pareto exponent. The reason for this popularity is arguably the simplicity and robustness of this method. Unfortunately, this procedure is strongly biased in small samples. We provide a simple practical remedy for this bias, and propose that, if one wants to use an OLS regression, one should use the Rank-1/2, and run log(Rank-1/2)=a-b log(Size). The shift of 1/2 is optimal, and reduces the bias to a leading order. The standard error on the Pareto exponent zeta is not the OLS standard error, but is asymptotically (2/n)^(1/2) zeta. Numerical results demonstrate the advantage of the proposed approach over the standard OLS estimation procedures and indicate that it performs well under dependent heavy-tailed processes exhibiting deviations from power laws. The estimation procedures considered are illustrated using an empirical application to Zipf's law for the U.S. city size distribution.

download in pdf format
   (384 K)

email paper

This paper is available as PDF (384 K) or via email.

An online appendix is available for this publication.

Acknowledgments

Machine-readable bibliographic record - MARC, RIS, BibTeX

Users who downloaded this paper also downloaded these:
Gabaix w14299 Power Laws in Economics and Finance
Rozenfeld, Rybski, Gabaix, and Makse w15409 The Area and Population of Cities: New Insights from a Different Perspective on Cities
Chen, Diebold, and Schorfheide w18078 A Markov-Switching Multi-Fractal Inter-Trade Duration Model, with Application to U.S. Equities
 
Publications
Activities
Meetings
Data
People
About

Support
National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email: info@nber.org

Contact Us