Aart Kraay, "Instrumental Variables Regression with Uncertain Exclusion Restrictions: A Bayesian Approach", Journal of Applied Econometrics, Vol. 27, No. 1, 2012, pp. 108-128. There are three data files corresponding to the three empirical examples in Section 4 of the paper. All three files are tab-delimited text files in DOS format. They are zipped in the file ak-data.zip. Unix/Linux users should use "unzip -a". AcemogluJohnsonRobinsonData.txt contains: name: Country name code: Country code loggdp: Logarithm of GDP per capita latitude: Absolute value of latitude expropriation risk: The expropriation risk measure of institutional quality used by AJR log settler mortality: The settler mortality variable used by AJR The data consist of a cross-section of 64 countries. The data are organized in columns corresponding to each variable. Variable names in first row, and country names are in the first column. FrankelRomerData.txt contains: ctry: Country name loggdp: Logarithm of GDP per capita actual trade share: exports plus imports divided by GDP constructed trade share: the geographically-predicted trade share used as instrument by FR log area: logarithm of area log population: logarithm of population The data consist of a cross-section of 150 countries. The data are organized in columns corresponding to each variable. Variable names in first row, and country names are in the first column. RajanZingalesData.txt contains: isic code: industry code country name: country name country number: numerical country code growth: growth of output in sector i in country j initial size: initial size of sector interaction: financial dependence interacted with financial development instrument: financial development interacted with effectiveness of legal system The data consist of a cross-section of 1067 country x industry observation. The data are organized in columns corresponding to each variable. Variable names in first row, industry codes are in the first column, and country names are in the second column.