Integrating and decomposing the sources of partisan bias: Brookes' method and the impact of redistricting in Great Britain

Partisan bias in British general elections

“…Moreover, the 1997 election does not appear as divergent from previous elections as Rossiter et al (1999) imply. There is also an important normative dimension to these figures.…”

“…There are two main families of methods for decomposing bias: the SoperDownloaded by [University of Chicago Library] at 20:33 25 December 2014 Rydon method (Soper and Rydon, 1958;Gudgin and Taylor, 1980;Curtice and Steed, 1992) and the Brookes method (Brookes, 1959;Mortimore, 1992;Johnston et al, 1999). Certain aspects of these methods will be considered in the next section; what matters for now is that uniform swing is assumed.…”

“…Gudgin and Taylor (1980) adapt the Soper-Rydon method to decompose the total amount of size bias into its constituent components (such as bias arising from differential turnout rates or differential numbers of electors). Johnston et al (1999) adapt the Brookes method for the same purpose. In this article, however, I simply report size bias as a whole, without decomposing it further.…”

Skrzywione podziały. Na tropie gerrymanderingu w polskich gminach

“…W połą-czeniu z innymi metodami (jak analiza symetrii partyjnej, zob. Grofman, King 2007;Johnston, Rossiter, Pattie 1999;Gilligan, Matsusaka 1999, Rycina 7. Zróżnicowanie procentu poparcia dla kandydatów komitetu inkumbenta i głównego pretendenta w Szczecinku w wyborach 2014 r., w zależności od wielkości okręgu Źródło: opracowanie własne na podstawie danych PKW.…”

From Gerry‐built to Purpose‐built: Drawing Electoral Boundaries for Unbiased Election Outcomes

“…Soper and Rydon (1958) then identified two drivers of bias: malapportionment and the geographic concentration of partisan support, which they called the 'differential concentration of majorities'. Johnston and others (Johnston et al 1999;Johnston et al 2001) have further developed measures for individual components of bias, differentiating between bias due to the number of votes lodged in each district (measured in turn as due to malapportionment, turnout differences or third party effects), and that due to the way that boundaries interact with the geographic distribution of partisan support (a single measure covers bias due to deliberate gerrymandering 6 or through following existing communities of interest).…”