Estimating party positions on immigration: Assessing the reliability and validity of different methods

ppqa_23_3.coverIt’s been in the making for a long time, but I’m happy to announce that Laura Morales and my paper on estimating party positions on immigration is now available from Party Politics. In the paper we provide a systematic assessment of various methods to position political parties on immigration based on their electoral platforms. We do this for Austria, Belgium, France, Ireland, the Netherlands, Spain, Switzerland, and the UK, between 1993 and 2013. There high levels of consistency between expert positioning, manual sentence-by-sentence coding, and manual checklist coding; and poor or inconsistent results with the CMP/MARPOR, Wordscores, Wordfish, and the dictionary approach. An often-neglected method – manual coding using checklists – offers resource efficiency with no loss in validity or reliability. Now there is really no excuse any more for using old CMP data and pretend that they really were about immigration… (with the new subcodes in the most recent codebook things will probably improve for the CMP/MARPOR positions).

We’ve started this as an internal project for the SOM project (hence 7 of the 8 countries), simply because we (thought we) needed party positions on immigration over time. Wary of the time it takes to manually code party manifestos, we tried a few methods. There are two more we have tried but not pursued to the same extent, namely using a dictionary of keywords and Wordfish estimates on the entire text of the party platforms (i.e. without manually selecting the parts of the manifestos that are about immigration). These are not ‘dead’ yet, but we need further tests to ensure we know what they measure.

There’s an 102-page supplement available from the journal’s website.

MIPEX in Switzerland: Some Clarifications

The Sonntags Zeitung was kind enough to comment on the recent release of MIPEX data in Switzerland. Unfortunately they seem more interested in the ranking than the details. The summary is roughly correct: it speaks of the immigrant integration policy framework.

The basic argument in the newspaper is logically incoherent, though: an index that places Switzerland behind France cannot be right, because France is struggling with integrating its immigrants (while there are no open conflicts in Switzerland). The editorial asks a series of questions about actual integration (“do we manage to integrate the many immigrants that arrive?”), not the policy framework. It refers to the high levels of youth unemployment among immigrants in France, but fails to connect this number to the youth unemployment in France more generally. Unemployment levels are generally much lower in Switzerland, but even this is beyond the point as it looks at outcomes and not the policy framework — this is what MIPEX does. Once immigrant policies and outcomes are conflated, it is easy to say that “there’s something wrong with these integration indicators” — well, that’s not what they are.

The main article provides a (factual) list of the main results and the positions of the MPG which are regarded rather critically. It would have been nice to distinguish between the indicators and the interpretations thereof (the ranking, “too restrictive”, “too many hurdles”), but to be fair, the way the MIPEX is usually presented this distinction is often blurred.