Welcome to The Log of Gravity page.
Reference (now with open access):
An early version of the paper can be found at CEP/LSE (and an even earlier version at Boston Fed).
In this page you can find the data set used in the paper, codes to extend some of the results in the paper, and other useful information on the implementation of the PPML estimator.
If you cannot find here the answer to your question about The Log of Gravity, please do not hesitate to contact the authors; we will be only too pleased to help.
The dataset (in xls and dta formats) is available here.
Testing competing models for non-negative data with many zeros:
We have developed a simple test to choose between competing models for non-negative data with many zeros. The paper is forthcoming in the Journal of Econometric Methods and is available here (an earlier version is here). Stata code to implement the test is available here; the code and data used in the applications are available here.
We have written a crude Stata command (recently updated) to estimate the IV version of PPML. This estimator was originally described in Windmeijer, F. and Santos Silva, J.M.C. (1997), "Estimation of Count Data Models with Endogenous regressors: An Application to Demand for Health Care," Journal of Applied Econometrics, 12(3), pp. 281-294. Here is an example of how to use the command. UPDATE: This estimator is now implemented in Stata 13; see the ivpoisson command and the add option (the default).
Here is a sample of the code to perform the test.
If you want to compute the R-squared for a model estimated by PPML, you can used the method implemented here.
PPML performance with many zeros:
Simulation evidence on the excellent performance of the PPML estimator when the data has many zeros can be found in this Economics Letters paper.
Be advised that there are several papers purporting to introduce estimators that improve on the PPML. While it is of course possible to find estimators that outperform PPML in specific conditions, to our knowledge all the newly proposed alternatives to PPML are either simply invalid, or valid only under implausibly strong distributional assumptions. Therefore we stand by the claim that PPML has all the characteristics needed to be the workhorse for the estimation of constant-elasticity models such as the gravity equation. If you believe to have evidence that another estimator generally outperforms PPML in this context please do let us know; we would be delighted to acknowledge that.
We have written a short reply to "The log of gravity revisited".
If you want to compute 'undertrading' and 'overtrading' after fixed-effects regressions with panel data, you need to obtain a set of residuals with zero mean. Here is how to do it.
Related work by other authors: Eight FAQ's & myths about the Log of Gravity
Eight FAQ's & myths about the Log of Gravity
According to Web-Counter, there have been unique visitors to this page since December 15, 2008
Last updated on 20 July 2014