Abstract
In a much cited 2008 article, Per Pettersson-Lidbom uses regression discontinuity to test for Swedish party effects on economic policies such as municipal taxation, spending and employment. We reassess the issue using the same estimator as Pettersson-Lidbom but new data on all factual coalitions, including minority coalitions as well as those previously deemed as undefined on a left and right wing political scale (constituting about 20 percent of the sample used in Pettersson-Lidbom’s study). This makes it possible to remove a systematic bias against centre-right coalitions in Pettersson-Lidbom’s study. We find that a majority of the previous findings stand, with sometimes even slightly stronger effects, but not as regards the proportional income tax rate and number of government employees per capita. Parties seem to matter for economic outcomes, but not always, and some parties more than others.