KNT Presentation Notes

Probability voter i votes for a i pr eua h qi i eub h

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ection of candidate j = A; B : X X i ;j EUji (h; qi ) = fV ;P (v ; p jH = h)U (v ; p ; qi ) + "i ;j p v Random utility setup with shocks "i ;j . Probability voter i votes for A: i Pr EUA (h; qi ) i EUB (h; qi ) We assume extreme value distribution: "i ;j i.i.d. F ("ij ) = exp ( e N XX dij ln Pr (Yi = j ) i =1 ln L( ) = e dij ln P i =1 = j N XX j l PP i ;j v fV ;P (v ;p jH =h )U (v ;p ;q i ) p e PP p i ;l v fV ;P (v ;p jH =h )U (v ;p ;q i ) Kendall, Nannicini & Trebbi (2014): “How Do Voters Respond to Information?” "ij ) Motivation Empirical model Experimental design Reduced-form results Model estimation Conclusion Non-Response Surveyed voters may choose not to disclose their vote Discarding this data may introduce bias if not ‘ missing completely at random’ We adopt approach of Ramalho & Smith (2012), modi…cation of choice-based (CB) sampling: assume probability of response is constant conditional on vote decision Estimate two additional response probabilities j for vote j = A; B N X X i e EU j (h ;qi ) ln L( ) = oi + dij ln j P EU...
View Full Document

This document was uploaded on 02/26/2014 for the course ECON 544 at UBC.

Ask a homework question - tutors are online