Party manifestos form the largest source of textual data for estimating party policy positions,
typically based on methods that assume that longer manifestos with more units of
text provide more confident estimates. Despite using them extensively for nearly three
decades, however, we know little to nothing about what explains why either the overall
length of manifestos or their scope of issue coverage varies so highly across parties, elections,
and contexts. Here, we critically test the notion that political context affects overall
length and manifesto content. We use multi-level modeling to predict manifesto length and
issue scope in a large number of coded party manifestos covering the post-war period. Our
findings indicate that manifesto length and the scope of issue coverage can be largely explained
by a combination of political variables related to party size, policy orientation, as
well as election-specific factors related to political competition and the timing of elections.
