Penelope A. Smith and Peter M Summers, "How Well Do Markov Switching Models Describe Actual Business Cycles? The Case of Synchronization", Journal of Applied Econometrics, Vol. 20, No. 2, 2005, pp. 253-274. The file "mssynch.zip" contains the data and programs for obtaining posterior distributions of the linear and the Markov switching VARs and for the concordance and correlation statistics between implied business cycle states. When unzipped the size of the folder is approximately 117 KB. The data file "newgdpd.txt" is in fixed column ASCII (DOS) format. The first column of this file is the date, with the subsequent columns containing Australian, US, UK, German, Canadian, Japanese quarterly, seasonally adjusted, real GDP data in that order. The sample period is 1961:I -2001:IV. In "newgdpd.txt" the effect of German reunification in 1991:I has been removed via a de-meaning regression of the growth rate of German real GDP on a dummy variable taking the value 1 in 1991:I. The original data value of quarterly growth rate of real German GDP obtained from DATASTREAM for this period was 13.44%. The adjusted quarterly growth rate of real German GDP in 1991:I is 0.68%. "MSVAR.G" is the main GAUSS program for estimating the posterior distributions parameters of the VAR and MSVAR. This code was modified from Richard Paap's "MARKVAR.PRG" code. This program makes use of the procedures "POWHALF.G," "PRIORHGT.G," "RNDINVW.G," and "MKERNEL.G," all of which were written by Richard Paap. Code for the Geweke-Hajivassiliou-Keane (GHK) simulator was obtained from V. Hajivassiliou's web page, http://econ.lse.ac.uk/staff/vassilis/index_own.html. StatTransfer7 was used to convert GAUSS (.dat) files containing Gibbs draws from the parameter vector to MATLAB (.mat) files. MATLAB programs for posterior inference on co-integrating relations are "abpost.m" and "cointpost.m". Data were simulated from the posterior distribution of the VAR and the MSVAR using the MATLAB files "simdata_var.m" and "simdata_msvar.m". The GAUSS program "dating_algorithm.g" is used to filter the simulated data with the BBQ algorithm. The MATLAB program "index_distn.m" calculates the degree of concordance and correlations between the business cycle states. Penelope Smith Melbourne Institute of Applied Economic and Social Research The University of Melbourne Parkville, Victoria, 3010 Australia Email: pasmit [AT] unimelb.edu.au