Hi,
I am estimating production functions using the Olley and Pakes method in Stata, specifically with the opreg command from Yasar, Raciborski, and Poi (2008). However, Stata is giving me the error message "an error occurred when bootstrap executed opreg - r(2000)". Now, r(2000) is the no observations error, but I checked that I did not eliminate all valid observations. Below first I'll describe the dataset in more detail, then I'll copy the code.
The dataset contains 1344 variables and 80782 observations, I don't think I can post any meaningful excerpt with dataex here. It contains ww2 US aircraft plants and the airplane models they produced, PlantByModelID gives a number to all plant-model pairs (e.g. the Lockheed-Burbank plant+their model A light bomber has ID number 145) (model name "Total" is total production), Model1 is the model name, time is measured in months for each plant-model pair (e.g. 1941m1, it's in %tm format), logOutputAMPR is a measure of output, Exitdummy is a dummy that is 1 if the plant isn't producing anymore, but produced in the previous period (unfortunately this dummy sometimes takes the value 1 several times per plant, as production was sometimes stopped and then restarted), the state variables measure capital and the age of the plant (logFloorTotal_Interp and PlantAge, respectively), logExpTotalCost measures investments, logHoursCU measures labour hours. Right now, I am only considering the total production in each plant in order to estimate the aircraft production function (model name is restricted to "Total"), so PlantByModelID is the plant identifier and Month is the time identifier in this panel. I am attaching the code I run below.
Does anyone know why Stata might throw the message "an error occurred when bootstrap executed opreg - r(2000)" here? Let me know if you need more information.
I am estimating production functions using the Olley and Pakes method in Stata, specifically with the opreg command from Yasar, Raciborski, and Poi (2008). However, Stata is giving me the error message "an error occurred when bootstrap executed opreg - r(2000)". Now, r(2000) is the no observations error, but I checked that I did not eliminate all valid observations. Below first I'll describe the dataset in more detail, then I'll copy the code.
The dataset contains 1344 variables and 80782 observations, I don't think I can post any meaningful excerpt with dataex here. It contains ww2 US aircraft plants and the airplane models they produced, PlantByModelID gives a number to all plant-model pairs (e.g. the Lockheed-Burbank plant+their model A light bomber has ID number 145) (model name "Total" is total production), Model1 is the model name, time is measured in months for each plant-model pair (e.g. 1941m1, it's in %tm format), logOutputAMPR is a measure of output, Exitdummy is a dummy that is 1 if the plant isn't producing anymore, but produced in the previous period (unfortunately this dummy sometimes takes the value 1 several times per plant, as production was sometimes stopped and then restarted), the state variables measure capital and the age of the plant (logFloorTotal_Interp and PlantAge, respectively), logExpTotalCost measures investments, logHoursCU measures labour hours. Right now, I am only considering the total production in each plant in order to estimate the aircraft production function (model name is restricted to "Total"), so PlantByModelID is the plant identifier and Month is the time identifier in this panel. I am attaching the code I run below.
Code:
xtset PlantByModelID Month opreg logOutputAMPR if Model1 == "Total", exit(Exitdummy) state(logFloorTotal_Interp PlantAge) proxy(logExpTotalCost) free(logHoursCU)
