Nonparametric structural analysis of discrete data: the quantile-based control function approach.
Doctoral thesis, UCL (University College London).
The first chapter is introduction and Chapter 2 proposes formal frameworks for identifiability and testability of structural features allowing for set identification. The results in Chapter 2 are used in other chapters. The second section of Chapter 3, Chapter 4 and Chapter 5 contain new results. Chapter 3 has two sections. The first section introduces the quantile-based control function approach (QCFA) proposed by Chesher (2003) to compare and contrast other results in Chapter 4 and 5. The second section contains new findings on the local endogeneity bias and testability of endogeneity. Chapter 4 assumes that the structural relations are differentiable and applies the QCFA to several models for discrete outcomes. Chapter 4 reports point identification results of partial derivatives with respect to a continuously varying endogenous variable. Chapter 5 relaxes differentiability assumptions and apply the QCFA with an ordered discrete endogeneous variable. The model in Chapter 5 set identifies partial differences of a nonseparable structural function.
|Title:||Nonparametric structural analysis of discrete data: the quantile-based control function approach|
|Open access status:||An open access version is available from UCL Discovery|
|UCL classification:||UCL > School of Arts and Social Sciences > Faculty of Social and Historical Sciences > Economics|
Archive Staff Only