Michael K. Salemi, "Estimating the Natural Rate of Unemployment and Testing the Natural Rate Hypothesis", Journal of Applied Econometrics, Vol. 14, No. 1, 1999, pp. 1-25. The data for the above paper are contained in the file data.file.mks. This file contains quarterly data for 1948:2 through 1990:1. Data for each quarter are stored as three 80-column-card images. Overall, the data set contains 504 card images (168 quarters x three card-images per quarter). Each quarter record contains 16 variables. The fortran format for the data is I5,2X,I2,G14.6/9X,5G14.6/9X,4G14.6. The variables and the symbols used for each in the paper are: 1. Year 2. Quarter 3. Unemployment rate U 4. Long duration unemployment rate Not used 5. Rate of change in manufacturing wage W 6. Inflation rate P 7. Rate of change in M1 M 8. Rate of change in price of fuels df 9. Index of help wanted ads Not used 10. Index of productivity Not used 11. Ratio of teen pop to adult pop Teen 12. Effective minimum wage EMW 13. Capital per capita Not used 14. Replacement rate RR 15. Probability that a randomly chosen worker qualifies for unemployment insurance Qprob 16. Price of fuels f The construction of each variable used in the paper is described in detail in the data appendix to the paper. The fortran code used to obtain the estimates reported in table 6 is contained in the code.file.mks. The estimation was done using the GQOPT package as described in the paper. The file includes fortran subroutines that are used in conjunction with GQOPT. The parameters that are read by the program at execution time are contained in parm.file.mks. All three files are zipped in the file mks.zip. They are in DOS format, so Unix users should use the -a option to remove extra carriage returns.