Misspecification in moment inequality models: back to moment equalities?

Abstract

Summary Consider the linear model where one is interested in learning about β given data on y and x and when y is interval measured; that is, we observe such that . Moment inequality procedures use the implication . As compared to least squares in the classical regression model, estimates obtained using an objective function based on these moment inequalities do not provide a clear approximation to the underlying unobserved conditional mean function. Most importantly, under misspecification, it is not unusual that no parameter β satisfies the previous inequalities for all values of x, and hence minima of an objective function based on these moment inequalities are typically tight. We construct set estimates for β in the linear model that have a clear interpretation when the model is misspecified. These sets are based on moment equality models. We illustrate these sets and compare them to estimates obtained using moment inequality‐based methods. In addition to the linear model with interval outcomes we also analyse the binary missing data model with a monotone instrument assumption (MIV), we find there that when this assumption is misspecified, bounds can still be non‐empty, and can differ from parameters obtained via maximum likelihood. We also examine a bivariate discrete game with multiple equilibria. In sum, misspecification in moment inequality models is of a different flavour than in moment equality models, and so care should be taken with (1) the_interpretation of the estimates and (2) the size of the ‘identified set’.

Set Identification, Moment Restrictions, and Inference

“…We will see that the partial identification technique is a way to protect oneself against this type of misspecification. However, the notion of partial identification lends itself less well to cases of misspecification of the family P θ,η (Ponomareva and Tamer, 2011).…”

Identification partielle : méthodes et conséquences pour les applications empiriques

“…Néanmoins la notion d'identification partielle se prête moins bien à cette extension (Ponomareva et Tamer, 2011) et c'est pourquoi la quasi-totalité de la littérature que nous passons en revue fait implicitement ou explicitement l'hypothèse de bonne spécification. Mauvaise traduction du difficile « sharp ».…”

Statistical modeling under partial identification: Distinguishing three types of identification regions in regression analysis with interval data

“…In the extended setting, however, two different types of identification regions have to be distinguished (cf. also [45]), called Sharp Marrow Region (SMR) and Sharp Collection Region (SCR), respectively here. The latter, arising from collecting all predictions based on all possible values compatible with the interval information, naturally is a superset of the first (cf.…”