We note that the existence of the maximum likelihood estimates in Poisson regression depends on the data configuration, and propose a strategy to identify the existence of the problem and to single out the regressors causing it.

We extend the simulation results in Santos Silva and Tenreyro (2006, The log of gravity, The Review of Economics and Statistics, 88, 641-658) by considering a novel data-generating process. Our results confirm that the Poisson pseudo-maximum likelihood estimator is generally well behaved, even...

A particular robust regression estimator has gained popularity among applied econometricians. We show that this estimator is inconsistent for the parameters of the conditional mean when the errors are skewed and heteroskedastic, and conclude that therefore its use cannot be generally recommended.

In this note, we argue that tests of overidentifying restrictions give little information on the validity of the moment conditions implied by the underlying economic model, and therefore are mute about the possibility of identifying the parameters of interest.

We highlight a subtle identification problem that afflicts models for non-negative data in which the conditional expectation is specified as the product of a logit and an exponential function. The results are illustrated with an empirical model for medical expenditures.