- Main
Essays on Identification
- Burkovskaya, Anastasia
- Advisor(s): Matzkin, Rosa L.;
- Zame, William R.
Abstract
Chapter 1 considers agents that aggregate several states of the world (i.e., the possible different realizations of an uncertain situation) into an event. They then compare bundles of consumption on these aggregated events, instead of having to consider bundles on each of the single possible realization. This simplifies the decision-making, since it decreases the number of distinguished realizations to consider. The collection of aggregated events of this type is called a subjective partition. In the case of Subjective Expected Utility, this kind of decision process does not affect agents’ choices. However, the presence of ambiguity (i.e., when agents do not know the probability associated with each possible realization of uncertain situations) changes their behavior dramatically: different subjective partitions lead to different preferences over bundles. Chapter 1 deals with preferences:
we provide axioms for a state aggregation model with ambiguity and show
under which conditions preferences imply uniqueness of the subjective partition.
Chapter 2 discusses choices obtained from the state aggregation model described in Chapter 1. We show identification of the subjective partition from
choices in a complete market setting. In addition, we demonstrate that the results can be applied to deductible insurance choices. Finally, we offer a set of revealed preference inequalities that allow for testing the model on a dataset.
Chapter 3 introduces a model of electoral choice that allows for derivation of
joint distribution of turnout and voter share from unobservable joint distribution
of costs of voting and preferences. Under a set of mild assumptions, we show
identification and provide non-parametric estimators of joint distribution of costs of voting and preferences over candidates from observable electoral data. All estimators are consistent and asymptotically normal.
Main Content
Enter the password to open this PDF file:
-
-
-
-
-
-
-
-
-
-
-
-
-
-