Data for Not so Harmless After All: The Fixed-Effects Model

Our results suggest that scholars ought to devote much more attention to modeling dynamics appropriately instead of relying on a default solution before they control for potentially omitted variables with constant effects using a fixed-effects specification.<br><br>Data record consists of 3 data files, in .do and .dta format, and a readme file. All Monte Carlo simulations in the paper have been executed with STATA 14 SE 64bit version and the following STATA ado-files installed: tabstatmat.ado , xml_tab.ado .

The data files are as follows: 1. nosoharmless_replication_final.do: STATA do-file: runs all MC simulations and generates all tables and figures in the article (tables 1, 2, 3a,b, 4a,b) and the online appendix (A1-A9), this file runs for ca. 60 hours on a machine similar to the one mentioned above. 2. results_mc_absbias_pa_final.dta: STATA data-file: contains all final MC results in a spreadsheet format 3. notsoharmless_replicaton_tables.do: STATA do-file: uses results_mc_absbias_pa_final.dta to generate all tables and figures in the article (tables 1, 2, 3a,b, 4a,b) and the online appendix (A1-A9), this file runs for a few seconds.