Premium
THE VARIABLE EFFECTS OF CHALLENGER SPENDING IN U.S. HOUSE ELECTIONS
Author(s) -
Berch Neil
Publication year - 1997
Publication title -
southeastern political review
Language(s) - English
Resource type - Journals
eISSN - 1747-1346
pISSN - 0730-2177
DOI - 10.1111/j.1747-1346.1997.tb00451.x
Subject(s) - margin (machine learning) , work (physics) , variable (mathematics) , baseline (sea) , economics , quality (philosophy) , political science , law , computer science , engineering , mechanical engineering , mathematical analysis , philosophy , mathematics , epistemology , machine learning
This paper extends the work on challenger spending in congressional elections to allow for the possibility that such spending has varying effects, depending upon circumstances. Using data from the 1988 and 1990 U.S. House elections, it begins with a baseline model for the incumbent's margin. Pairs of dummy variables and interaction terms were then added to test several hypotheses. The results provide some evidence to suggest that challenger spending is less effective if the challenger faced a toughprimary. Contrary to previous work, an advantage was found for repeat challengers. Also contrary to previous work, the results do not suggest that high‐quality challengers spend their money more efficiently than do other challengers.