The materials here will replicate the main results of the paper "Too much of a good thing? Estimating the trade-off between voter turnout and the representativeness of election results when studying turnout shocks.", being Figure 2 and Tables 1 and 3. The enclosed files (almost all of which are Stata-files) will create the dataset used for the analysis and the main figures and tables of the paper. The code begins with files from input sources. That is, all steps of the process, including data cleaning, are performed in this replication. Note that in the paper we also use data from a LISS panel survey called "Wiv Referendum 2018: Third measurement". This data is not included in the files and is not required for running the do-file. The agency that ran this survey prefers it when researchers directly download this (opensource) data from their website: https://www.dataarchive.lissdata.nl/study-units/view/815 (doi: 10.17026/dans-xcr-n8az) <strong style="background-color:rgb(255, 238, 204);">Derived From: ...