Research output: Thesis › Doctoral Thesis
Research output: Thesis › Doctoral Thesis
}
TY - BOOK
T1 - Essays on Stochastic Frontier Models
AU - Jonuzaj, Mariol
PY - 2023/7
Y1 - 2023/7
N2 - This thesis consists of three independent chapters. The first chapter “Bayesian Inference in Dynamic Panel Stochastic Frontier Models” proposes a new stochastic frontier model which accounts for the intertemporal production behaviour. The conceptualization is based on the notion that firms face production adjustment costs in the short run due to the presence ofquasi-fixed inputs. Consequently, this sluggish adjustment of the entire production process will create a dependency between the current and past production state. To capture this dynamic process, this chapter utilizes the traditional partial adjustment mechanism. The mechanism delivers a dynamic specification and allows factor inputs and inefficiency shocks to have anintertemporal effect on the production process. Moreover, the model allows heterogeneous adjustment speeds and input elasticities across the production units. Model inference is based on Bayesian MCMC techniques with data-augmentation. We illustrate the new model in an empirical application where we estimate the productivity and efficiency growth of the Egyptian private manufacturing sector during the early 90’s.In a similar vein, the second chapter, “Dynamic Panel Stochastic Frontier Modelswith Inefficiency Effects”, deals with dynamic panel frontier models where inefficiency effects can be a function of exogenous environmental variables. This chapter builds upon advancements in the field and utilizes parametric cumulative distribution functions to specify technical efficiency. The proposed model allows the presence of fixed effects and time-varying inefficiencies.Model estimation is based on the Generalized Method of Moments (GMM) approach, where various forms of input endogeneity can be effectively addressed.Last, the third chapter “A simple method for modelling the energy efficiencyrebound effects with an application to energy demand frontiers” proposes a newsimple method for estimating the energy inefficiency rebound effects. Model estimation is based on a two-stage approach. In the first stage, we argue in estimating a reduced form stochastic frontier model with country-specific inefficiency heteroscedastic effects. In the second stage, the energy efficiency rebound effects can be obtained effectively using moment-matching methodssuch as the GMM approach. We apply the proposed model on aggregate energy frontiers where we estimate the energy efficiency and the corresponding rebound effects for a balanced panel of OECD economies. The empirical results suggest an overall upward trend of energy efficiency scores. The energy rebound effects range from 28% to 92%, indicating that energy efficiency actions could have a limited impact on achieving environmental objectives.
AB - This thesis consists of three independent chapters. The first chapter “Bayesian Inference in Dynamic Panel Stochastic Frontier Models” proposes a new stochastic frontier model which accounts for the intertemporal production behaviour. The conceptualization is based on the notion that firms face production adjustment costs in the short run due to the presence ofquasi-fixed inputs. Consequently, this sluggish adjustment of the entire production process will create a dependency between the current and past production state. To capture this dynamic process, this chapter utilizes the traditional partial adjustment mechanism. The mechanism delivers a dynamic specification and allows factor inputs and inefficiency shocks to have anintertemporal effect on the production process. Moreover, the model allows heterogeneous adjustment speeds and input elasticities across the production units. Model inference is based on Bayesian MCMC techniques with data-augmentation. We illustrate the new model in an empirical application where we estimate the productivity and efficiency growth of the Egyptian private manufacturing sector during the early 90’s.In a similar vein, the second chapter, “Dynamic Panel Stochastic Frontier Modelswith Inefficiency Effects”, deals with dynamic panel frontier models where inefficiency effects can be a function of exogenous environmental variables. This chapter builds upon advancements in the field and utilizes parametric cumulative distribution functions to specify technical efficiency. The proposed model allows the presence of fixed effects and time-varying inefficiencies.Model estimation is based on the Generalized Method of Moments (GMM) approach, where various forms of input endogeneity can be effectively addressed.Last, the third chapter “A simple method for modelling the energy efficiencyrebound effects with an application to energy demand frontiers” proposes a newsimple method for estimating the energy inefficiency rebound effects. Model estimation is based on a two-stage approach. In the first stage, we argue in estimating a reduced form stochastic frontier model with country-specific inefficiency heteroscedastic effects. In the second stage, the energy efficiency rebound effects can be obtained effectively using moment-matching methodssuch as the GMM approach. We apply the proposed model on aggregate energy frontiers where we estimate the energy efficiency and the corresponding rebound effects for a balanced panel of OECD economies. The empirical results suggest an overall upward trend of energy efficiency scores. The energy rebound effects range from 28% to 92%, indicating that energy efficiency actions could have a limited impact on achieving environmental objectives.
U2 - 10.17635/lancaster/thesis/2023
DO - 10.17635/lancaster/thesis/2023
M3 - Doctoral Thesis
PB - Lancaster University
ER -