Search Results
Title: Three Essays on Applied Economics
Resulting in 1 citation.
| 1. |
Katayama, Hajime |
|
Three Essays on Applied Economics Ph.D. Dissertation, The Pennsylvania State University, 2003. DAI-A 64/09, p. 3416, Mar 2004 Cohort(s): NLSY97 Publisher: UMI - University Microfilms, Bell and Howell Information and Learning Keyword(s): Crime; Heterogeneity; Labor Economics; Modeling, Multilevel; Modeling, Probit This thesis consists of three essays. The first essay is joint work with Kala Krisha and Susumu Imai. Using data from the National Youth Survey, we examine the relationship of current criminal activity with past criminal activity, past arrests, and other variables. We estimate an ordered probit model, allowing for unobserved heterogeneity. We find that criminal types and non-criminal types behave very differently. An increase in arrests raises current criminal activity only for non-criminal types, while an increase in criminal experience raises current criminal activity for both types. For both types, arrests rise and then fall with age with a peak around age 18. The age crime profile also has this shape for non-criminal types, but for criminal types, it rises with age, suggesting lower apprehension rates for criminal types. The second essay looks at the pattern of young males' criminal activity, using data from the National Longitudinal Survey of Youth 97. A multivariate Tobit model with random effects is estimated to investigate whether current criminal acts of one type affect future criminal activity in the same or other categories. The results indicate such relationships indeed exist. In particular, I find evidence that minor crimes such as vandalism and minor theft are a stepping stone to more serious crimes. This suggests discouraging these will reduce future crime. The third essay is joint work with Shihua Lu and James Tybout. We develop an approach to measuring firms' performances. We assume firms' costs and revenues reflect a Bertrand-Nash equilibrium in a differentiated product industry. Given the demand system parameters, this allows us to impute each firm's unobserved marginal costs and product quality from its observed revenues and costs. Assuming that marginal costs and product quality indices follow vector autoregressive (VAR) processes, we jointly estimate the demand system parameters and VAR parameters using Bayesian techniques. Applying our methodology to panel data on Colombian pulp and paper plants, we find that conventional productivity measures are not closely related to quality measures and are nearly orthogonal to consumer surplus measures, suggesting that they may be poor characterizations of producer performance from a social welfare standpoint. |
|
|
Bibliography Citation
Katayama, Hajime. Three Essays on Applied Economics. Ph.D. Dissertation, The Pennsylvania State University, 2003. DAI-A 64/09, p. 3416, Mar 2004. |