Replication package for: "Global and local technical changes: A new decomposition of the Malmquist Productivity Index using virtual units"
Description
This replication package contains the files required for replicating the results presented in the paper “Global and local technical changes: A new decomposition of the Malmquist Productivity Index using virtual units” co-authored by Juan Aparicio and Daniel Santín and accepted for publication by January 2024 in Economic Modelling. The package contains: - The well-known database of 42 Swedish pharmacies between 1980 and 1989 previously analyzed in the seminal paper of Färe et al. (1992). They showed how to implement and decompose the Malmquist productivity index (MPI) using data envelopment analysis (DEA). This decomposition of total factor productivity changes consists of two terms; the catching-up effect or efficiency change and the frontier-shift effect or technical change (TC). This dataset was also used in other seminal papers as in Balk and Althin (1996), and Althin (2001). - R code with the data in Table 1 and the steps for doing the numerical analysis and achieving the results showed in Table 2 and Fig. 4. It also allows to obtain results in Table 3 changing K. - R code with the steps for doing the numerical analysis and obtaining GTC and LTC reproducing Tables 4 and 5 and Fig. 5. - A README file basic instructions for running the R scripts and the definition of variables (for more details see the papers above). References: Althin, R. (2001). Measurement of productivity changes: Two Malmquist index approaches. Journal of Productivity Analysis, 16(2), 107-128. Balk, B. M., & Althin, R. (1996). A new, transitive productivity index. Journal of Productivity Analysis, 7(1), 19-27. Färe, R., Grosskopf, S., Lindgren, B., & Roos, P. (1992). Productivity changes in Swedish pharmacies 1980-1989: A non-parametric Malmquist approach. Journal of Productivity Analysis, 3(1), 85-101.
Files
Steps to reproduce
The general guidelines for using the R scripts are: 1) Open any of the two R scripts in RStudio. 2) The excel file pharmacy.xlsx is composed by ten spreadsheets, one for each year. 3) Replace in all the instructions "read_excel" your own paths to open the pharmacy.xlsx database from where you have saved it on your computer. For example, change the original path d80 <- read_excel("C:/Daniel/EM/pharmacy.xlsx", sheet = "R80") by d80 <- read_excel("C:/Manuela/Myproject/ReplicationEM/pharmacy.xlsx", sheet = "R80") 4) Run the R code