Post by northbruin40 on Jun 13, 2021 11:02:38 GMT -8
Some analysis on how California has trended this Century centering on the sum of Congressional districts.
CalCongressTrends.pdf (176.51 KB)
The graph looks at the sum of party votes (in Democrat - Republican) for California's 53 Congressional districts. It shows both the simple sum (solid black curve) and my adjusted difference (dashed black curve). The reason for the adjustment is not every district general election is a classic Democrat vs. Republican contest. It can be two members of the same party, or in a few cases a candidate running unopposed. That tends to warp the results and exaggerates (in my opinion) the net difference. So I applied some corrections for the dashed curve (for instance Presidential results can give an estimate of the party leans in a district).
There are really two big jumps over the past nearly 20 years - from 2004 to 2008 and from 2014 to 2018. As for add-ons - a simple formula can reasonable predict the Presidential election results in California for 2008 to 2020 - based upon Party registration difference + a constant (the graph shows +11%).
As to why the adjustment should be applied, look at these selected district results for 2014-2020. South of the Tehachapi Mountains, the district changes often really have been that huge - as indicated by the dashed black line. If you don't due the adjustment, the curve is biased by too many Democrat vs. Democrat general election contests (by the way, yes there have been Republican vs. Republican general cases too), as a consequence of the jungle primaries.
CalDistricts.pdf (207.58 KB)
CalCongressTrends.pdf (176.51 KB)
The graph looks at the sum of party votes (in Democrat - Republican) for California's 53 Congressional districts. It shows both the simple sum (solid black curve) and my adjusted difference (dashed black curve). The reason for the adjustment is not every district general election is a classic Democrat vs. Republican contest. It can be two members of the same party, or in a few cases a candidate running unopposed. That tends to warp the results and exaggerates (in my opinion) the net difference. So I applied some corrections for the dashed curve (for instance Presidential results can give an estimate of the party leans in a district).
There are really two big jumps over the past nearly 20 years - from 2004 to 2008 and from 2014 to 2018. As for add-ons - a simple formula can reasonable predict the Presidential election results in California for 2008 to 2020 - based upon Party registration difference + a constant (the graph shows +11%).
As to why the adjustment should be applied, look at these selected district results for 2014-2020. South of the Tehachapi Mountains, the district changes often really have been that huge - as indicated by the dashed black line. If you don't due the adjustment, the curve is biased by too many Democrat vs. Democrat general election contests (by the way, yes there have been Republican vs. Republican general cases too), as a consequence of the jungle primaries.
CalDistricts.pdf (207.58 KB)
