May 302016

Privately, I have referred to this piece as The Un-Dead Article, the Paper That Is Never Going Be Published, The Cursed Manuscript, or simply as It-Which-Must-Not-Be-Named. But you know, it’s the problem child we love the most. So: Our article “Political interest furthers partisanship in England, Scotland, and Wales” is finally out! If you don’t have a subscription (please check this first), here is an ungated link that works for a very limited number of visitors (please consider your fellow impoverished HE institution). And if everything fails, here is my pre-print version: Political Interest Furthers Partisanship in England, Scotland, and Wales.

United Kingdom Flag Map

The article argues, first, that the extant literature on party identification in the UK underestimates levels of identification, because it lumps together respondents from three different party systems (England, Scotland, and Wales). Second, we take the very useful model proposed by Clarke and his associates, who treat party identification as a latent class, and make a minor adjustment by adding political interest as an explanatory variable. As it turns out, political interest makes identification more likely. This is more in line with classic ideas about party identification than with “revisionist” critiques of the Michigan model, and with current models of political cognition. Moreover, it suggests that political interest renders affective ties more powerful in stabilizing themselves.

Mar 062015

Last week, I showed you that partisan dealignment in the Western federal states of Germany has slowed down considerably, and essentially came to a standstill during the last decade. But what are the mechanisms behind this trend? I think that one part of the puzzle is the changing role of formal education.

More Formal Education, Less Party Identification?

In the quaint olde world that was Western Germany, formal education and party identification were supposed to be negatively correlated. One reason for that is simple: Voters with higher levels of formal education were less likely to belong to the working class, and less likely to be frequent churchgoers, hence less likely to belong to the respective core constituencies of the two major parties. A second, more intriguing reason is cognitive mobilisation (Dalton 1984). In this view, party identification serves as a heuristic that reduces the cognitive costs of processing political information. Educated, more interested voters, however, face lower cognitive costs and have therefore fewer incentives to develop partisan attachments. Thus, the rising level of educational attainment should undermine partisanship.

Formal Education and Party Identification: A Changing Relationship

More recently, doubts have been raised about the merits of cognitive mobilisation theory, as there is growing evidence that politically sophisticated voters are more, not less likely to be partisans (Albright 2009, Dassonneville et al. 2012). One possible reason for this is the changing role of educational attainment in Germany. Three or even two decades ago, the vast majority of voters had spent ten years or less in school, but for the younger generations, the Abitur is slowly becoming the norm. Low levels of attainment are now increasingly associated with political and social marginalisation.

Dealignment in Germany by Formal Education (Model-Based)

Dealignment in Germany by Formal Education (Model-Based)

At any rate, education has an increasingly positive effect in my individual-level model of partisanship in the Western states for the 1977-2012 period. Controlling for age group, time and a lot of nifty interactions, it would seem that in recent years, dealignment has mostly been confined to the lower-education strata, while in the growing upper educational echelons, the downward trend in partisanship has petered out for the time being, slowing down the overall decline.


Feb 242015

Almost a decade ago, I published an article with a cutesy title on the decline of party identification in Germany, of which I am inordinately proud. The main message of this piece was that party identification in Germany has not collapsed, but is rather declining at the glacial rate of 0.7 percentage points per year, give or take. Here is the relevant graph:

Party Identification in West Germany, 1977-2002

For a more recent project, I have extended the time-series to cover the whole 1977-2012 period, right up to the begin of the 2013 federal election campaign. As it turns out, de-alignment in the West has come to a virtual halt during the last decade – see here:

Partisan Dealignment in West Germany

Decline of party identification in West Germany, 1977-2012

If you think that this is still too noisy, have a look at this trajectory, which is derived from a binary logistic model that regresses identifications on time, age, education, and campaign effects. More on this soon – stay tuned.

Estimated overall levels of partisanship in West Germany, 1977-2012 (adjusted predictions at representative values (APR)

Estimated overall levels of partisanship in West Germany, 1977-2012 (adjusted predictions at representative values (APR)

Jun 072011

Five years ago, I published a paper on the apparently inevitable decline of party identifications in Germany. The somewhat cutesy title of the piece is Dead Men Walking. It is based on the ‘Politbarometer’ series of monthly polls going back all the way to the late 1970s, and in my humble opinion, it is a rather neat application of the “analysing repeated surveys” approach. One of my main findings is that on average, the share of party identifiers declines at a rate of about 0.7 percentage points per year. Recently, I re-ran my scripts on a new data set that extends the old series all through the naughties. As you can see, party ID in Germany is not exactly alive and kicking, but the rate of decline has fallen considerably over the last decade. As one wise man once observed, the core problem with predictions is that they are about the future.

Party Identification in Germany (% identifiers)

Sep 232010

Someone asked me for the syntax/data required to replicate my old Electoral Studies piece on party identification in Germany 1977-2002. Its slightly preposterous title not withstanding, as of today it holds a proud 18th rank in Electoral Studies’ list of its current “Top 25 hottest articles” (bringing Web 2.0 and the sciences together was always bound to end up in disaster). So in the very unlikely event that you have been holding your breath for this, breath out: I finally got around to de-clutter my old files and uploaded the replication information to my dataverse:

Kai Arzheimer, 2010, “Replication data for: Dead Men Walking? Party Identity in Germany 1977-2002”, hdl:1902.1/15091 UNF:5:AkxnvIlPgabqb2zsZ39k1A==