Kristian Behrens, Cem Ertur, and Wilfried Koch, "'Dual' Gravity: Using Spatial Econometrics to Control for Multilateral Resistance", Journal of Applied Econometrics, Vol. 27, No. 5, 2012, pp. 773-794. The data are provided in two formats. The file data.txt is an ASCII file in DOS format that is zipped in bek-data.zip. The file bek-data.xls is an Excel file that is zipped in bek-data-excel.zip. Unix/Linux users should use "unzip -a" on the former but must *not* use the "-a" option on the latter. The variables in data.txt are described in estimgrav.m and in the Excel file bek-data.xls. Various Matlab .m files, all of which are ASCII files in DOS format, are zipped in bek-programs.zip. Unix/Linux users should use "unzip -a". In order to estimate the model, we use Matlab routines of James Lesage found at: http://www.spatial-econometrics.com/. In order to replicate our estimation, it is necessary to download their routines and use the joint .m file: estimgrav.m It is also necessary to keep all the .m files in the same directory. The files sargrav.m and sacgrav.m include information matrices. sacgrav.m corresponds to the sarar model in the paper. sarma.m and LVsarma.m allow homogenous coefficients estimation of our model whereas sarmadouble.m and LVsarmadouble.m allow heterogenous coefficients estimation of our model.