Cornell University
Library
Cornell UniversityLibrary

eCommons

Help
Log In(current)
  1. Home
  2. Cornell University Graduate School
  3. Cornell Theses and Dissertations
  4. Essays In Labor Economics And Synthetic Data Methods

Essays In Labor Economics And Synthetic Data Methods

File(s)
Gittings, Robert.pdf (840.06 KB)
Permanent Link(s)
https://hdl.handle.net/1813/14039
Collections
Cornell Theses and Dissertations
Author
Gittings, Robert
Abstract

Three topics are investigated in these chapters: the causes and consequences of lateral job mobility within firms, the impact of incentives on human behavior in the context of capital punishment and deterrence, and the development of new synthetic data methods for confidentiality protection of public use data. The extent and importance of lateral mobility is not well-established in economics and Chapter 1 contributes new and important findings to the literature. Using a panel of more than 500 firms and 48,000 white-collar workers, I find relatively high rates of lateral mobility, that this mobility is statistically different from other transitions, and that the compensation growth associated with lateral mobility is economically meaningful. I also investigate the relationships between worker performance, compensation growth and job mobility. Even when controlling for productivity differences, significant earnings growth occurs directly through the change in jobs. The results provide some evidence that the observed lateral mobility may be the result of job rotation. In light of continued debate of whether capital punishment deters crime, Chapter 2 revisits my previous work on this issue and shows that the deterrence results hold under alternative measurements of key variables, multiple statistical specifications and subsets of the data. Chapter 3 develops methodology that solves the need for statistical agencies to suppress certain data items because releasing those cells to the public yields a risk of exposing someone's personal information. I show that the synthetic data adequately protect the confidential data and are superior in terms of its analytical validity.

Date Issued
2009-10-14T20:05:06Z
Type
dissertation or thesis

Site Statistics | Help

About eCommons | Policies | Terms of use | Contact Us

copyright © 2002-2026 Cornell University Library | Privacy | Web Accessibility Assistance