Daniel J. Henderson, Christopher F. Parmeter, and R. Robert Russell, "Modes, Weighted Modes and Calibrated Modes: Evidence of Clustering using Modality Tests," Journal of Applied Econometrics, Vol. 23, No. 5, 2008, pp. 607-638. There are 24 separate text files, all of them ASCII files in DOS format. They are all zipped in the file hpr-data.zip. In addition, there is a data appendix in pdf format as well as the actual GAUSS files which can be used to replicate the results. The latter are zipped in hpr-progs.zip. Data description There are four types of files to be aware of. The files ending with a w refer to specific datasets with a column for the population. These files are year specific. Files beginning with GDPL refer to the six gdp series used in the paper for the smaller dataset (the L refers to a longer time-series; 1960-2000, with fewer countries; 97). Files beginning with GDPS refer to the six gdp series used in the paper for the larger dataset (shorter time-series; 1970-2000, with more countries; 138). The order in these data sets is a country code, the country's population (if applicable), then cgdp, rgdpl, rgdpch, rgdpeqa, rgdpwok, and rgdptt. For example, the file GDPL00w.txt contains a country code, the population of the country and the six measures of income (cgdp, rgdpl, rgdpch, rgdpeqa, rgdpwok and rgdptt), for the group of 97 countries in the year 2000. The file GDPLu.txt contains the same country code for the 97 countries in the first column but is ordered as cgdp, rgdpl, rgdpch, rgdpeqa, rgdpwok and rgdptt in 1960, cgdp, rgdpl, rgdpch, rgdpeqa, rgdpw1ok and rgdptt in 1970, ... , cgdp, rgdpl, rgdpch, rgdpeqa, rgdpwok and rgdptt in 2000. The files beginning with FE refer to the datasets constructed in the spirit of Feyrer (2003). Note here that the S and L have a different meaning than with the GDP files. Here L refers to the group of 77 countries from 1960-2000, while the S refers to the 93 countries over the 1970-2000 period. Those ending in a w have a column for the country populations while those that do not (FEL.txt and FES.txt) contain data over the entire sample period. For example, the file FES00w.txt contains a country code, the country population, rgdpch, KY, MPK, H-25, H-15, TFP-25, and TFP15 for the 93 countries in the year 2000 used in the analysis for the 1960-2000 period. The FESu.txt folder has the same order but data for all years. Specifically, the dataset contains a country code, followed by rgdpch, KY, MPK, H-25, H-15, TFP-25, and TFP15 in 1970, rgdpch, KY, MPK, H-25, H-15, TFP-25, and TFP15 in 1980, ... , rgdpch, KY, MPK, H-25, H-15, TFP-25, and TFP15 in 2000. The file Bianchi.txt contains the original Bianchi (1997) data. The file country.txt refers to the country codes created for the datasets. Code description To run the GAUSS code, the library file modal.lcg will need to be placed in the library (lib) folder of your main GAUSS folder, and modal.src will need to be placed in the source (src) folder of your main GAUSS folder. The file bump.g runs both the uncalibrated and calibrated versions of the Silverman (1981) modality test, while the file caldip.g runs the calibrated version of the dip test of Hartigan and Hartigan (1985).