panel.production: Randomly generated panel of production data

In snfa: Smooth Non-Parametric Frontier Analysis

Description

A dataset for illustrating technical and efficiency changes using smooth non-parametric frontiers.

Usage

panel.production

Format

A data frame with 200 observations of six variables.

Firm

Firm identifier

Year

Year of observation

X.1

Input 1

X.2

Input 2

X.3

Input 3

y

Output

Details

Generated with the following code:

set.seed(100)

num.firms <- 20
num.inputs <- 3
num.years <- 10

beta <- runif(num.inputs, 0, 1)
TFP.trend = 0.25
TFP <- cumsum(rnorm(num.years)) + TFP.trend * (1:num.years)

sd.measurement <- 0.05
sd.inefficiency <- 0.01

f <- function(X){
  return(TFP + X 
}
gen.firm.data <- function(i){
  X = matrix(runif(num.years * num.inputs, 1, 10), ncol = num.inputs)
  y = f(X) +
    rnorm(num.years, sd = sd.measurement) -
    abs(rnorm(num.years, sd = sd.inefficiency))
  firm.df <- data.frame(Firm = i,
                        Year = 1:num.years,
                        X = exp(X),
                        y = exp(y))
}

panel.production = Reduce(rbind, lapply(1:num.firms, gen.firm.data))
panel.production$Firm = as.factor(panel.production$Firm)

snfa documentation built on May 1, 2019, 9:16 p.m.