| OCR Text |
Show 96 everyone and anyone who may want to verify, conduct their own examination, and/or run the documents through software (to compare to human-annotated versus computerannotated results) (Merz, Regel, and Lewandowski 2016). This seems to be a concern that MARPOR has been taking steps to actively address. While MARPOR may not be perfect, it continually strives to improve its dataset, and, for those of us interested in time series and cross sectional analyses of parties and elections, it is superior to expert surveys, like Chapel Hill (CHES), because it contains more time points. MARPOR starts in the postWWII era, while the Chapel Hill Expert Survey has five waves, which begin in 1999. In resuming the discussion of variables, in addition to considering the independent variables individually, I also created nine interaction terms: two pair unemployment with the MCCP niche measures, four pair GDP with the environmental measures, and three pair MCCP or environmental niches' dedication to their "big" issue (the percent of the niche party platform dedicated to the niche issue) with the corresponding sum of the top three mainstream parties' platform on MCCP, MCCP opposite, and environmental issues. These interaction terms serve two purposes. First of all, it will allow me to test Hypotheses 14 and 15, based on arguments in the literature that certain conditions, like inflows of foreigners, only matter in the presence of others, like unemployment. In other words, if taken individually, either or neither may appear to have a statistically significant impact; only when they interact are their effects apparent. Secondly, interacting the strategies of niche and mainstream parties is a central component of the strategic interaction model. Parties do not operate in a vacuum, detached from other parties, so to more fully understand how parties, in this case niche and mainstream, shape electoral |
