This paper will be primarily descriptive and employ automated quantitative text analysis to provide a first series of summary indicators for the data on parliamentary questions and speeches collected for the PATHWAYS project. The empirical base will be the texts of parliamentary questions for written answer tabled in by all MPs during the last completed session of the national parliaments covered by the PATHWAYS project. Our research design is based on a dual comparison (a) between countries and (b) between CIO MPs and a matching sample of other MPs within each country. Since quantitative analyses of speeches and other forms of substantive representation are extremely rare and patchy for European parliaments, even the descriptive character of the paper will cover completely new ground both in relation to individual countries as well as with regard to the comparisons. Given the differences in language and semantics, the paper will start by identifying the methodological problems arising from the quantitative analysis of legislative speeches across languages and institutional contexts. Subsequently we will present the results of (a) country-specific topic models, (b) of sentiment analyses comparing CIO and non-CIO MPs in relation to policy issues pertaining to immigrants and citizens of immigrant origin and (c) analyses of influential ‘separating words’ in parliamentary questions, i. e. terms that are highly predictive for particular populations and sub-populations. The findings will be presented by using standard techniques of visualizing differences between legislators and countries. Where the number of MPs and size of the text corpus allows, these comparisons will account for differences across stages of the parliamentary cycle and the career cycle as well as party affiliation of MPs with a CIO background. In addition the paper will look at the extent to which policy areas are framed in terms of migration and integration.
