Description
|
We study a unique quasi-experiment in Austria, where compulsory voting laws are changed across Austria’s nine states at different times. Analyzing state and national elections from 1949-2010, we show that compulsory voting laws with weakly enforced fines increase turnout by roughly 10 percentage points. However, we find no evidence that this change in turnout affected government spending patterns (in levels or composition) or electoral outcomes. Individual-level data on turnout and political preferences suggest these results occur because the impacts of compulsory voting on turnout are larger among those who are non-partisan, who have low interest in politics, and who are uninformed. (2017-01-04)
|
Notes
| Readme file to replicate Tables and Figures in: Hoffman, León and Lombardi. “Compulsory Voting, Turnout, and Government Spending: Evidence from Austria”. Journal of Public Economics, Vol. 145: 103–115. January 2017. The zipped file contains all do files, raw data sets and working data sets. In order to replicate the results from our paper, you need to paste the contents of this folder in your computer and change the base directory in the 7th line of the STATA do-file “Master.do”. The entire analysis can be run from “Master.do”. The folder “Raw Data Files” contains the following raw data sets: - “1980-2012_expenditures.dta” holds the information on nominal spending for each of the nine Austrian states, broken down into ten spending categories for 1980-2012, as provided by the Austrian Statistical Agency. - “cpi_index.dta” contains the Austrian Consumer Price Index (base year 2010), obtained from http://stats.oecd.org - “population_unemployment_new2.dta” has the yearly population and unemployment rates of all Austrian states, provided by the Austrian Statistical Agency. - “Election-Parliament.dta”, “Election-President.dta” and “Election-State.dta” contains voting data from the parliamentary, presidential and state elections held in Austria in 1945-2010, broken down by state. This data was obtained from the Austrian Federal Ministry of the Interior’s yearbooks. The zip folder contains a do-file generating the working data sets. In particular, “Database_Construction_Elections_Expenditures.do” constructs the “Elections.dta” data set in lines 8-278, and the “Expenditures.dta” data set in lines 284-430. You will need to install the “carryforward” command in STATA by typing “ssc install carryforward”. We construct our Tables and Figures from three working datasets: - “Elections.dta” contains the data on turnout and election results needed to construct Figure 2, Panel A of Table 1, and Tables 3, 5, A.3, A.8 (columns 1 and 2), A.9, A.11, and A.12 - “Expenditures.dta” contains the data on state-level expenditure and additional variables needed for Figure 2, Panel B of Table 1, and Tables 4, A.5, A.6, A.7, A.8 (columns 3-6), and A.10. - “ASS.dta” holds the relevant variables constructed from the 1986 and 2003 waves of the ASS, necessary for Tables 2 and 6. Due to data confidentiality reasons, we do not disclose the original files from the Austrian Social Survey, but we do disclose the Do File we used to construct it (“Database_Construction_ASS.do”). Figures 2 and all Tables in the paper are built from the “Tables and Figures.do” do file. In order to calculate the wild-bootstrap p-values, it is first necessary to obtain the “cgmreg.ado” and “cgmwildboot.ado” ado-files from Judson Caskey’s website (https://sites.google.com/site/judsoncaskey/data). You also need to install the “unique” command in STATA by typing “ssc install unique”. The wild-bootstrap p-values featured in the paper were obtained from versions 12 or 13 of STATA (the random number generator differs slightly in STATA 14). If you are working in STATA 14, type “version 13.0”. To create Figure 2 you will have to install the grc1leg package in STATA by typing “ssc install grc1leg”. |